From owner-reliable_computing [at] interval [dot] louisiana.edu Tue Feb 1 10:50:04 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j11Go3oD027071 for ; Tue, 1 Feb 2005 10:50:03 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j11Go3Jc027070 for reliable_computing-outgoing; Tue, 1 Feb 2005 10:50:03 -0600 (CST) Received: from imap.univie.ac.at (mailbox-lmtp.univie.ac.at [131.130.1.27]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j11Gnrls027066 for ; Tue, 1 Feb 2005 10:49:59 -0600 (CST) Received: from [131.130.16.23] (theseus.mat.univie.ac.at [131.130.16.23]) by imap.univie.ac.at (8.12.10/8.12.10) with ESMTP id j11GniUW247612; Tue, 1 Feb 2005 17:49:47 +0100 Message-ID: <41FFB2FB.4050503 [at] univie [dot] ac.at> Date: Tue, 01 Feb 2005 17:48:59 +0100 From: Arnold Neumaier Organization: University of Vienna User-Agent: Mozilla Thunderbird 0.9 (X11/20041127) X-Accept-Language: en-us, en MIME-Version: 1.0 To: interval Subject: Interval arithmetic and the SIAM 100-digit challenge Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-DCC-ZID-Univie-Metrics: mx9.univie.ac.at 4248; Body=2 Fuz1=2 Fuz2=2 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk I just noticed that the review of the book F. Bornemann et al., The SIAM 100-digit challenge, SIAM, Philadelphia 2004 in the MathSciNet http://www.ams.org/msnmain?fmt=doc&fn=105&id=2076374&l=20&pg3=IID&r=1&s3=254928&v3=Bornemann%2C%20Folkmar refers to interval arithmetic for verifying the accuracy of the results. ''The authors second objective in each chapter is to validate the correctness of each calculated digit. This objective is achieved through various combinations of carefully designed computer experiments, a posteriori error estimates and computer-assisted proofs based on interval arithmetic.'' Arnold Neumaier From owner-reliable_computing [at] interval [dot] louisiana.edu Tue Feb 1 11:12:11 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j11HCBwY027203 for ; Tue, 1 Feb 2005 11:12:11 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j11HCA03027202 for reliable_computing-outgoing; Tue, 1 Feb 2005 11:12:10 -0600 (CST) Received: from smtp2.rz.tu-harburg.de (smtp2.rz.tu-harburg.de [134.28.205.13]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j11HC1VJ027198 for ; Tue, 1 Feb 2005 11:12:07 -0600 (CST) Received: from mail2.rz.tu-harburg.de (mail2.rz.tu-harburg.de [134.28.202.179]) by smtp2.rz.tu-harburg.de (8.13.1/8.13.1) with ESMTP id j11HC0aa018352 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Tue, 1 Feb 2005 18:12:00 +0100 Received: from oem-9n90y5vdt8b (d187111.adsl.hansenet.de [80.171.187.111]) (user=ti3sr mech=LOGIN bits=0) by mail2.rz.tu-harburg.de (8.13.1/8.13.1) with ESMTP id j11HBxUi012448 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NO); Tue, 1 Feb 2005 18:11:59 +0100 To: "Arnold Neumaier" , interval Subject: Re: Interval arithmetic and the SIAM 100-digit challenge References: <41FFB2FB.4050503 [at] univie [dot] ac.at> Message-ID: Date: Tue, 01 Feb 2005 18:11:55 -0100 From: "Siegfried M. Rump" Organization: Technical University Hamburg-Harburg Content-Type: text/plain; format=flowed; delsp=yes; charset=iso-8859-15 MIME-Version: 1.0 Content-Transfer-Encoding: 8bit In-Reply-To: <41FFB2FB.4050503 [at] univie [dot] ac.at> User-Agent: Opera M2/7.54 (Win32, build 3869) X-Scanned-By: TUHH Rechenzentrum content checker on 134.28.205.13 X-Scanned-By: TUHH Rechenzentrum content checker on 134.28.202.179 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Yes, one of the authors wrote me some time ago that five of the ten problems were solved using INTLAB. Best wishes Siegfried M. Rump On Tue, 01 Feb 2005 17:48:59 +0100, Arnold Neumaier wrote: > I just noticed that the review of the book > F. Bornemann et al., > The SIAM 100-digit challenge, > SIAM, Philadelphia 2004 > in the MathSciNet > http://www.ams.org/msnmain?fmt=doc&fn=105&id=2076374&l=20&pg3=IID&r=1&s3=254928&v3=Bornemann%2C%20Folkmar > refers to interval arithmetic for verifying the accuracy of the results. > > ''The authors second objective in each chapter is to validate the > correctness of each calculated digit. This objective is achieved through > various combinations of carefully designed computer experiments, a > posteriori error estimates and computer-assisted proofs based on > interval arithmetic.'' > > > Arnold Neumaier > -- ================================================= Prof. Dr. Siegfried M. Rump Arbeitsbereich Informatik III Technische Universität Hamburg-Harburg Schwarzenbergstr. 95 21071 Hamburg Germany phone +49 40 42878 3027 secr. +49 40 42878 3227 fax +49 40 42878 2489 From owner-reliable_computing [at] interval [dot] louisiana.edu Tue Feb 1 18:32:54 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j120Wsiv027831 for ; Tue, 1 Feb 2005 18:32:54 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j120WswF027830 for reliable_computing-outgoing; Tue, 1 Feb 2005 18:32:54 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j120WqTo027826 for ; Tue, 1 Feb 2005 18:32:52 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j120WpiP027825 for reliable_computing [at] interval [dot] louisiana.edu; Tue, 1 Feb 2005 18:32:51 -0600 (CST) Received: from postfix3-2.free.fr (postfix3-2.free.fr [213.228.0.169]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j0S9Pjxc018202 for ; Fri, 28 Jan 2005 03:25:51 -0600 (CST) Received: from Tartaros (longchamp44-3-82-224-250-165.fbx.proxad.net [82.224.250.165]) by postfix3-2.free.fr (Postfix) with ESMTP id 036C8C24B; Fri, 28 Jan 2005 10:25:32 +0100 (CET) From: "Christophe Jermann" To: "Christophe Jermann" Subject: Postdoctoral position in Nantes, France Date: Fri, 28 Jan 2005 10:25:31 +0100 MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Mailer: Microsoft Office Outlook, Build 11.0.6353 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 Thread-Index: AcUFG09tET/E1MVVR6u+NucyvlbdxQ== Message-Id: <20050128092532.036C8C24B@postfix3-2.free.fr> Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk [apologies if you receive multiple copies of this message] POSTDOCTORAL POSITION in COMPUTER SCIENCE and INTERVAL MATHEMATICS University of Nantes, LINA laboratory, CoCoA Team ----- The CoCoA (Continuous Constraints and Applications) team of the LINA (Computer Science Laboratory of Nantes Atlantique) laboratory invites applications for a postdoctoral position. The appointment will be for one year, starting between September 2005 and December 2005. The CoCoA team will support candidates into applying to fellowships and grants programs. For this purpose, interested candidates should send their application the sooner. Context ------- The CoCoA team consists of 15 researchers who aim at solving general first-order formulas over integers and reals, involving non-linear constraints, uncertainty, preferences on the data and efficiency criteria. The base tools we use are interval analysis, constraint propagation (CP), symbolic computation and tree-search methods. Constraint programming provides a declarative and natural way of formulating problems as constraint satisfaction problems (CSPs) by stating the requirements (constraints) that must be fulfilled by the solutions. The CoCoA team is especially interested in numerical CSPs which involve constraints over real numbers. Such CSPs appear in industrial applications like conceptual design, computer aided design, robotics and molecular biology, which are the main applications the CoCoA team is tackling. Objectives ---------- The postdoctoral fellow will be involved into one of the ongoing research projects of the team, contribute to the development of the ELISA platform for constraint programming and optimization, and pursue researches in one of the following directions (depending on her/his abilities and desires): * hybridization of solving/optimization techniques (local/global, numeric/symbolic, ...) * design of solving/optimization methods for composite problems (hierarchical conjunctive/disjunctive models, mixed problems) * definition of systems and languages for constraint programming and optimization * study of a given application field (design, bioinformatics, ...) Candidate profile ----------------- Interested candidates should send an application letter including detailed CV (with a list of publications and a description of research interests) to the contacts below. The candidate must hold a recent PhD (within 5 years) at the appointment time. Applicants must have sound publications and expertise in one or more of the following areas (in a broad sense): * Programming languages and constraint programming * Optimization techniques (interval, convex, local, global, ...) * Cooperative problem solving Environment ------------ The city of Nantes is ideally located, only two hours from Paris by TGV (high-speed train, 20 shuttles per day) and about two hours by plane from most European capitals. Gateway to the Brittany ports, just a step away from major tourist sites such as Mont Saint-Michel, the Puy du Fou, the Futuroscope or the "Chateaux de la Loire", Nantes is also only 50km away from the renown coasts of Brittany and beaches of Vendee. The LINA laboratory offers a friendly working environment and excellent computational facilities. Contacts -------- Laurent Granvilliers or Christophe Jermann Laboratoire LINA - University of Nantes 2 rue de la Houssiniere BP 92208 F-44322 Nantes CEDEX France Phone: +33 251 125 851 or +33 251 125 840 Fax: +33 251 125 812 email: {laurent.granvilliers, christophe.jermann}@univ-nantes.fr http://www.sciences.univ-nantes.fr/lina/en/research/teams/COCOA/index.html From owner-reliable_computing [at] interval [dot] louisiana.edu Thu Feb 3 18:39:35 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j140dZeG000952 for ; Thu, 3 Feb 2005 18:39:35 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j140dZ13000951 for reliable_computing-outgoing; Thu, 3 Feb 2005 18:39:35 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j140dQYj000944 for ; Thu, 3 Feb 2005 18:39:32 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j140d3x28149; Thu, 3 Feb 2005 17:39:03 -0700 (MST) Message-Id: <200502040039.j140d3x28149 [at] cs [dot] utep.edu> Date: Thu, 3 Feb 2005 17:39:02 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: a book of potential interest to interval researchers: forwarding To: reliable_computing [at] interval [dot] louisiana.edu, interval [at] cs [dot] utep.edu Cc: Michael.Oberguggenberger [at] uibk [dot] ac.at MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: 8Djyq/w7zFLESJDm2PgDCg== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk From: "Michael Oberguggenberger" ... Dear colleagues, It is my pleasure to bring your attention to the edited monograph W. Fellin, H. Lessmann, M. Oberguggenberger, R. Vieider: Analyzing Uncertainty in Civil Engineering. Springer-Verlag, Berlin 2005. You can find the table of contents at http://www.cs.utep.edu/interval-comp/fellin.pdf Interval-related papers are in this book in as much as random sets and fuzzy sets are employed, and this implies using evaluations of functions on intervals. In my review article, intervals as models of uncertainty are discussed as well; in addition, there is a paper on sensitivity analysis. Thus the book might be of interest to the interval community. On behalf of the editors, I thank you for your interest. With best regards, Michael Oberguggenberger Institut fuer Technische mathematik, Geometrie und Bauinformatik Universitaet Innsbruck Technikerstr. 13 A - 6020 Innsbruck, Austria Tel ++43 (0)512 507 6824 Fax ++43 (0)512 507 2941 E-Mail michael.oberguggenberger [at] uibk [dot] ac.at From owner-reliable_computing [at] interval [dot] louisiana.edu Fri Feb 4 23:28:54 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j155Sr1j001928 for ; Fri, 4 Feb 2005 23:28:53 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j155SrHO001927 for reliable_computing-outgoing; Fri, 4 Feb 2005 23:28:53 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j155SgGL001923 for ; Fri, 4 Feb 2005 23:28:48 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j155SYH06877; Fri, 4 Feb 2005 22:28:34 -0700 (MST) Message-Id: <200502050528.j155SYH06877 [at] cs [dot] utep.edu> Date: Fri, 4 Feb 2005 22:28:36 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: interval talks welcome at an El Paso 2006 conference To: reliable_computing [at] interval [dot] louisiana.edu, interval [at] cs [dot] utep.edu Cc: solin [at] math [dot] utep.edu, rafi.muhanna [at] gtrep [dot] gatech.edu, Robert.Mullen [at] case [dot] edu MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: QsgFWChbVEPTbqBHgz88wg== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear Friends, FEM CONFERENCE: IN BRIEF In the second week of December 2006, El Paso, Texas will host an international workshop on Finite Element Methods (FEM) and their applications to Science and Engineering. Several renowned researchers are planning to come, some funding has already been secured, and the organizers are actively applying for additional support. This workshop is organized by an enthusiastic inter-disciplinary team: * Pavel Solin (http://www.math.utep.edu/Faculty/solin/) is an applied mathematician specializing in higher order finite element methods; he is the author of a research monograph Higher-Order Finite Element Methods (Chapman & Hall/CRC Press, 2003) * Jack Chessa (http://www.dmc.utep.edu/chessa) is a professor of mechanical engineering department, a specialist in finite element methods for non-linear problems such as fracture, shear band and localization phenomena as well as fluid structure interaction and multi-phase flow, and, more generally, problems that deal with moving interfaces and discontinuities. INTERVAL TALKS ARE VERY WELCOME Both Pavel and Jack have heard a lot of good things about interval techniques in FEM as techniques that provide guaranteed error bounds, they are very much interested to learn more about these techniques and they strongly believe that the interaction between leading FEM researchers and interval researchers interested in FEM applications will be very beneficial to both sides. POTENTIAL COMPLEMENT TO REC WORKSHOPS This aspect of the planned workshop makes is a nice complement to the Second International Workshop on Reliable Engineering Computing (REC) that Rafi Muhanna and Robert Mullen plan to organize in Savannah, Georgia, in February 2006 - after the success of their first workshop in September 2004: * the REC bring together researchers who use validated techniques in engineering (not necessarily in FEM), where we can exchange our ideas and techniques and advertise our successes to the larger engineering community * on the other hand, the planned FEM workshop will give us an opportunity to make our ideas, techniques, and results more known in the general engineering computation community, especially among those who use FEM methods. PLEASE EXPRESS YOUR INTEREST ASAP If you are potentially interested in coming, please let Pavel know ASAP by email to solin [at] math [dot] utep.edu (copy to me at vladik [at] cs [dot] utep.edu). By March 1, 2005, he would like to have a preliminary list. The more impressive the resulting list, the larger potential for collaboration and applications, the better chances for getting additional funding (for some of the grants to support 2006 conferences, early March 2005 is the deadline). By March 1, the workshop dates will be decided exactly. INTERVAL TUTORIAL SOLICITED In view of the collaboration potential, in addition to the research interval-related talks of interest to the FEM community, the organizers would like to include a description of interval techniques with an emphasis on FEM as a tutorial and/or one of the keynote talks. They will try their best to fund potential keynote speakers. During several interval-related conferences and at the REC meeting in Savannah, I have heard many interesting talks and tutorials that, in my opinion and in the opinion of other participants, are very clear so that they can be rather easily understood by the engineering-related audience and very enthusiastic so that they can enthuse participants into using these verified (reliable) techniques. If you are interested in presenting such a tutorial and/or keynote talk or if you would like to suggest someone who, in your opinion, would be the best person for this job, please do not hesitate to email Pavel (copy to me) about these possibilities. VENUE The workshop will be held either on campus or at the new Hilton hotel that Hilton is currently building on our campus (it is scheduled to open in February 2006). In any case, a block of rooms will be reserved at the Hilton hotel for the out-of-town guests. The local tradition is that in early December, a Season of Light is officially opened, with thousands of holiday lights decorating the trees and the Bhutanese-style buildings on campus. These lights are on every evening in December, so you will have an opportunity to enjoy them. From owner-reliable_computing [at] interval [dot] louisiana.edu Sat Feb 5 11:58:28 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15HwRrI003670 for ; Sat, 5 Feb 2005 11:58:27 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j15HwRLn003669 for reliable_computing-outgoing; Sat, 5 Feb 2005 11:58:27 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15HwIAR003665 for ; Sat, 5 Feb 2005 11:58:24 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j15HwEi09476; Sat, 5 Feb 2005 10:58:14 -0700 (MST) Message-Id: <200502051758.j15HwEi09476 [at] cs [dot] utep.edu> Date: Sat, 5 Feb 2005 10:58:12 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: FW: calculators To: reliable_computing [at] interval [dot] louisiana.edu Cc: pjwilkinson [at] pasadena [dot] edu MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: yYYk0bJ9UaAUABPK6y/rQQ== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Re need to teach kids (and adults) about accuracy and maybe intervals. And maybe in favbor of interval calculators. ------------- Begin Forwarded Message ------------- Date: Sat, 5 Feb 2005 09:46:52 -0800 From: "Paul Wilkinson, Ph. D." Subject: FW: calculators To: SIGCSE-MEMBERS [at] LISTSERV [dot] ACM.ORG I spent a long time teaching mathematics ... I have found that if you take the calculator away from the student, they cannot add, subtract, multiply, or divide, let alone doing fractions. An example that I have had arguments with students over is (2/3)^4 * (3/2)^4. The answer we all know is 1, but they argue their calculator says the answer is not one, but something, depending upon the calculator brand, like 1.00000001. ... From my observations, it appears as if all they know how to do is punch buttons when they get out of high school when calculators are used in the classroom. Paul J. Wilkinson, Ph. D. Chair Department of Information Systems and Computer Science Pasadena City College Pasadena, California ------------- End Forwarded Message ------------- From owner-reliable_computing [at] interval [dot] louisiana.edu Sat Feb 5 12:07:37 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15I7aJe003788 for ; Sat, 5 Feb 2005 12:07:36 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j15I7aAS003787 for reliable_computing-outgoing; Sat, 5 Feb 2005 12:07:36 -0600 (CST) Received: from ms-smtp-02-eri0.ohiordc.rr.com (ms-smtp-02-smtplb.ohiordc.rr.com [65.24.5.136]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15I7Ro7003783 for ; Sat, 5 Feb 2005 12:07:33 -0600 (CST) Received: from Moore (wor226111.columbus.rr.com [204.210.226.111]) by ms-smtp-02-eri0.ohiordc.rr.com (8.12.10/8.12.7) with SMTP id j15I7EXV004870; Sat, 5 Feb 2005 13:07:14 -0500 (EST) Message-ID: <000701c50bad$912cd790$1702a8c0@Moore> From: "Ray Moore" To: "Vladik Kreinovich" , Cc: References: <200502051758.j15HwEi09476 [at] cs [dot] utep.edu> Subject: Re: calculators Date: Sat, 5 Feb 2005 13:07:33 -0500 MIME-Version: 1.0 Content-Type: text/plain; format=flowed; charset="iso-8859-1"; reply-type=original Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.2180 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 X-Virus-Scanned: Symantec AntiVirus Scan Engine Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk No, I think the point is that we still need to teach kids mental arithmetic when it is that easy. Ray Moore ----- Original Message ----- From: "Vladik Kreinovich" To: Cc: Sent: Saturday, February 05, 2005 12:58 PM Subject: FW: calculators > Re need to teach kids (and adults) about accuracy and maybe intervals. And > maybe in favbor of interval calculators. > > ------------- Begin Forwarded Message ------------- > > Date: Sat, 5 Feb 2005 09:46:52 -0800 > From: "Paul Wilkinson, Ph. D." > Subject: FW: calculators > To: SIGCSE-MEMBERS [at] LISTSERV [dot] ACM.ORG > > I spent a long time teaching > mathematics > ... > > I have found that if you take the calculator away from the student, > they cannot add, subtract, multiply, or divide, let alone doing fractions. > > An example that I have had arguments with students over is (2/3)^4 * > (3/2)^4. The answer we all know is 1, but they argue their calculator says > the answer is not one, but something, depending upon the calculator brand, > like 1.00000001. > > ... > > From my observations, it appears as if all they know how to do is punch > buttons when they get out of high school when calculators are used in the > classroom. > > Paul J. Wilkinson, Ph. D. > Chair > Department of Information Systems and Computer Science > Pasadena City College > Pasadena, California > > > ------------- End Forwarded Message ------------- > > > From owner-reliable_computing [at] interval [dot] louisiana.edu Sat Feb 5 12:17:13 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15IHDTS003906 for ; Sat, 5 Feb 2005 12:17:13 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j15IHCwE003905 for reliable_computing-outgoing; Sat, 5 Feb 2005 12:17:12 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15IH4ld003901 for ; Sat, 5 Feb 2005 12:17:10 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j15IGxF09561; Sat, 5 Feb 2005 11:16:59 -0700 (MST) Message-Id: <200502051816.j15IGxF09561 [at] cs [dot] utep.edu> Date: Sat, 5 Feb 2005 11:16:58 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: Re: calculators To: vladik [at] cs [dot] utep.edu, reliable_computing [at] interval [dot] louisiana.edu, rmoore17 [at] columbus [dot] rr.com Cc: pjwilkinson [at] pasadena [dot] edu MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: ou5pGfWUswYyj1huM+3pHQ== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear Ray, >From your reply, I realize that I did not make my point right. I agree with what you said, the need to teach mental arithmetic was the main point (and a valid point) of the posting, a point that I did not emphasize because it is of general interest, not specific to our interval area. What I wanted to say that is that on top of this point, adding uncertainty to calculators - whether in explanations in the class or explicitly in hardware -- would further help students. In the example given by Dr. Wilkinson, yes, mental arithmetic would help, but if they try square root of 2 then they need to know that the result is also not exact. >From the viewpoint of a kid, a weight measured by a scale is more accurate that a weight guessed by a human being; a velocity measured by a radar is more accurate than estimated visually. A natural kid's conclusion is that whatever calcuator produces shouldbe more accurate than what we do by hand. This is one of the sources of confusion. Honestly, when I balance by checkbook by hand, I trust myself less (and make mistakes more often) then if I use the corresponding software. Calculator is capable of correct (guaranteed) computations, but we need to program and modify it accordingly, or at least explain. I think the need felt by educators may be a good opportunity for us to educate people more in uncertainty issues. Vladik > From: "Ray Moore" > To: "Vladik Kreinovich" , > Cc: > Subject: Re: calculators > Date: Sat, 5 Feb 2005 13:07:33 -0500 > MIME-Version: 1.0 > Content-Transfer-Encoding: 7bit > X-Priority: 3 > X-MSMail-Priority: Normal > X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 > X-Virus-Scanned: Symantec AntiVirus Scan Engine > > No, I think the point is that we still need to teach kids mental arithmetic > when it is that easy. > > Ray Moore > > ----- Original Message ----- > From: "Vladik Kreinovich" > To: > Cc: > Sent: Saturday, February 05, 2005 12:58 PM > Subject: FW: calculators > > > > Re need to teach kids (and adults) about accuracy and maybe intervals. And > > maybe in favbor of interval calculators. > > > > ------------- Begin Forwarded Message ------------- > > > > Date: Sat, 5 Feb 2005 09:46:52 -0800 > > From: "Paul Wilkinson, Ph. D." > > Subject: FW: calculators > > To: SIGCSE-MEMBERS [at] LISTSERV [dot] ACM.ORG > > > > I spent a long time teaching > > mathematics > > ... > > > > I have found that if you take the calculator away from the student, > > they cannot add, subtract, multiply, or divide, let alone doing fractions. > > > > An example that I have had arguments with students over is (2/3)^4 * > > (3/2)^4. The answer we all know is 1, but they argue their calculator says > > the answer is not one, but something, depending upon the calculator brand, > > like 1.00000001. > > > > ... > > > > From my observations, it appears as if all they know how to do is punch > > buttons when they get out of high school when calculators are used in the > > classroom. > > > > Paul J. Wilkinson, Ph. D. > > Chair > > Department of Information Systems and Computer Science > > Pasadena City College > > Pasadena, California > > > > > > ------------- End Forwarded Message ------------- > > > > > > From owner-reliable_computing [at] interval [dot] louisiana.edu Sat Feb 5 13:29:16 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15JTGZo004116 for ; Sat, 5 Feb 2005 13:29:16 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j15JTFdv004115 for reliable_computing-outgoing; Sat, 5 Feb 2005 13:29:15 -0600 (CST) Received: from relay.pair.com (relay00.pair.com [209.68.1.20]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with SMTP id j15JT6SQ004111 for ; Sat, 5 Feb 2005 13:29:12 -0600 (CST) Received: (qmail 81823 invoked from network); 5 Feb 2005 19:29:05 -0000 Received: from unknown (HELO ?192.168.1.100?) (unknown) by unknown with SMTP; 5 Feb 2005 19:29:05 -0000 X-pair-Authenticated: 24.244.69.242 Mime-Version: 1.0 X-Sender: jat711k [at] www [dot] peda.com Message-Id: In-Reply-To: <200502051816.j15IGxF09561 [at] cs [dot] utep.edu> References: <200502051816.j15IGxF09561 [at] cs [dot] utep.edu> Date: Sat, 5 Feb 2005 11:28:04 -0800 To: reliable_computing [at] interval [dot] louisiana.edu From: Jeff Tupper Subject: Re: calculators Cc: pjwilkinson [at] pasadena [dot] edu Content-Type: text/plain; charset="us-ascii" ; format="flowed" Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Back in 2000, I worked (through my company) with Hewlett-Packard on the Xpander calculator project. The calculator, which could have introduced some interval concepts to a larger audience (I was responsible for the reliable graphing portion), was cancelled---after prototypes were built and demonstrated---as upper management at HP unexpectedly decided to put nearly all calculator development on hold. The large manufacturers have some interest in interval methods but want to see demand for such methods from their customers. Jeff From owner-reliable_computing [at] interval [dot] louisiana.edu Sat Feb 5 13:49:46 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15JnkEB004243 for ; Sat, 5 Feb 2005 13:49:46 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j15JnjP4004242 for reliable_computing-outgoing; Sat, 5 Feb 2005 13:49:45 -0600 (CST) Received: from ms-smtp-01-eri0.ohiordc.rr.com (ms-smtp-01-smtplb.ohiordc.rr.com [65.24.5.135]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15Jnb3b004238 for ; Sat, 5 Feb 2005 13:49:42 -0600 (CST) Received: from Moore (wor226111.columbus.rr.com [204.210.226.111]) by ms-smtp-01-eri0.ohiordc.rr.com (8.12.10/8.12.7) with SMTP id j15JnUHu002885; Sat, 5 Feb 2005 14:49:34 -0500 (EST) Message-ID: <001201c50bbb$dcc2a1e0$1702a8c0@Moore> From: "Ray Moore" To: "Jeff Tupper" , Cc: References: <200502051816.j15IGxF09561 [at] cs [dot] utep.edu> Subject: Re: calculators Date: Sat, 5 Feb 2005 14:49:50 -0500 MIME-Version: 1.0 Content-Type: text/plain; format=flowed; charset="iso-8859-1"; reply-type=response Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.2180 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 X-Virus-Scanned: Symantec AntiVirus Scan Engine Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Jeff and all, And how could such "demand" be demonstrated ? Ray Moore ----- Original Message ----- From: "Jeff Tupper" To: Cc: Sent: Saturday, February 05, 2005 2:28 PM Subject: Re: calculators > Back in 2000, I worked (through my company) with Hewlett-Packard on > the Xpander calculator project. The calculator, which could have > introduced some interval concepts to a larger audience (I was > responsible for the reliable graphing portion), was cancelled---after > prototypes were built and demonstrated---as upper management at HP > unexpectedly decided to put nearly all calculator development on > hold. The large manufacturers have some interest in interval methods > but want to see demand for such methods from their customers. > > Jeff > From owner-reliable_computing [at] interval [dot] louisiana.edu Sat Feb 5 14:04:39 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15K4ccV004361 for ; Sat, 5 Feb 2005 14:04:39 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j15K4ca7004360 for reliable_computing-outgoing; Sat, 5 Feb 2005 14:04:38 -0600 (CST) Received: from ms-smtp-01-eri0.ohiordc.rr.com (ms-smtp-01-smtplb.ohiordc.rr.com [65.24.5.135]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15K4TB0004356 for ; Sat, 5 Feb 2005 14:04:35 -0600 (CST) Received: from Moore (wor226111.columbus.rr.com [204.210.226.111]) by ms-smtp-01-eri0.ohiordc.rr.com (8.12.10/8.12.7) with SMTP id j15K45Hu017850; Sat, 5 Feb 2005 15:04:05 -0500 (EST) Message-ID: <002001c50bbd$e43bdca0$1702a8c0@Moore> From: "Ray Moore" To: "Vladik Kreinovich" , Cc: References: <200502051816.j15IGxF09561 [at] cs [dot] utep.edu> Subject: Re: calculators Date: Sat, 5 Feb 2005 15:04:24 -0500 MIME-Version: 1.0 Content-Type: text/plain; format=flowed; charset="iso-8859-1"; reply-type=original Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.2180 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 X-Virus-Scanned: Symantec AntiVirus Scan Engine Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Vladik, > What I wanted to say that is that on top of this point, adding uncertainty > to > calculators - whether in explanations in the class or explicitly in > hardware -- > would further help students. HOW would it help them ? > In the example given by Dr. Wilkinson, yes, mental arithmetic would help, > but > if they try square root of 2 then they need to know that the result is > also not > exact. WHY would they "need to know that the result is also not exact" ? > From the viewpoint of a kid, a weight measured by a scale is more accurate > that > a weight guessed by a human being; a velocity measured by a radar is more > accurate than estimated visually. A natural kid's conclusion is that > whatever > calcuator produces shouldbe more accurate than what we do by hand. This is > one > of the sources of confusion. HOW does this cause confusion ? > Honestly, when I balance by checkbook by hand, I trust myself less (and > make > mistakes more often) then if I use the corresponding software. A pocket calculator is perfectly capable, if used correctly, and without any intervals, of balancing your checkbook correctly; and so you are by hand if you take the time. > Calculator is capable of correct (guaranteed) computations, but we need to > program and modify it accordingly, or at least explain. That is not necessary for just checkbook calculations. Better examples are needed to generate "demand" from prospective.customers. Ray Moore From owner-reliable_computing [at] interval [dot] louisiana.edu Sat Feb 5 14:26:03 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15KQ3wp004495 for ; Sat, 5 Feb 2005 14:26:03 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j15KQ3d0004494 for reliable_computing-outgoing; Sat, 5 Feb 2005 14:26:03 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15KPpt4004490 for ; Sat, 5 Feb 2005 14:26:00 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j15KPkD10036; Sat, 5 Feb 2005 13:25:46 -0700 (MST) Message-Id: <200502052025.j15KPkD10036 [at] cs [dot] utep.edu> Date: Sat, 5 Feb 2005 13:25:45 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: Re: calculators To: vladik [at] cs [dot] utep.edu, reliable_computing [at] interval [dot] louisiana.edu, rmoore17 [at] columbus [dot] rr.com Cc: pjwilkinson [at] pasadena [dot] edu MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: ZhWkHH1U9RKpikV4+njlmg== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear Ray, Dear participants of the mailing list, I seem to be not in my clearest today, my apologies. Of course, I did not mean that interval calculators somehow help in balancing a checkbook. I apologize for creating this false impression. Let me try to state my case clearer. Let me describe a scenario how current calculators can cause confusion and how interval calculators can help. * First, a student balances a checkbook by hand and then again by using a calculator. In many such cases, the two results are somewhat different. So, the student checks again and again, and find the mistake in what he did by hand. This happens several times, with different examples. As a result, the student acquires the following rule: if a calculator produces one result and computations by hand produce a different result, then, most probably, a calculator is right and calculations by hand are wrong. * Now, a student computes the square root of 2 and then squares the resulting number. If he or she does it "by hand" (i.e., without using a calculator, in this case, by writing expressions on a sheet of paper, by using reasoning), this student should get the exact same value 2. The student is well aware that he sometimes makes mistakes. So, to be on the safe side, this students checks his or her result on the calculator. The result is not exactly 2, but something like 2.00000001. The student already knows, from earlier painful experience, that when the calculator and his manual results differ, there is a serious probability that the calculator is right. (Even when the results of, say, checkbook balancing were differing by 1 cent between what the student got by hand and what the student got by using a calculator, still the most usual case was a student's mistake, not the calculator's mistake.) This is where the student is confused: instead of trusting the results of his own work, he trusts the calculator instead, and this is a wrong occasion to trust the calculator: because with real numbers, calculators are only approximate devices. * How would an interval calculator help: to check the value 2 that the student got by arguing on a sheet of paper, the student runs the same computations on a calculator. As a result, the student gets an interval [1.99999999,2.000001] that clearly contains his result 2. So, instead of being confused (as with the traditional calculator), the student is reassured that his or her original sheet-of-paper calculations were correct. I hope I made my case clearer now. Vladik P.S. I agree that balancing a checkbook is not the right example for convincing the general audience, I am simply trying to explain why I forwarded the original email to the interval mailing list in the first place. > From: "Ray Moore" > To: "Vladik Kreinovich" , > Cc: > Subject: Re: calculators > Date: Sat, 5 Feb 2005 15:04:24 -0500 > MIME-Version: 1.0 > Content-Transfer-Encoding: 7bit > X-Priority: 3 > X-MSMail-Priority: Normal > X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 > X-Virus-Scanned: Symantec AntiVirus Scan Engine > > Vladik, > > > What I wanted to say that is that on top of this point, adding uncertainty > > to > > calculators - whether in explanations in the class or explicitly in > > hardware -- > > would further help students. > > HOW would it help them ? > > > In the example given by Dr. Wilkinson, yes, mental arithmetic would help, > > but > > if they try square root of 2 then they need to know that the result is > > also not > > exact. > > WHY would they "need to know that the result is also not exact" ? > > > From the viewpoint of a kid, a weight measured by a scale is more accurate > > that > > a weight guessed by a human being; a velocity measured by a radar is more > > accurate than estimated visually. A natural kid's conclusion is that > > whatever > > calcuator produces shouldbe more accurate than what we do by hand. This is > > one > > of the sources of confusion. > > HOW does this cause confusion ? > > > Honestly, when I balance by checkbook by hand, I trust myself less (and > > make > > mistakes more often) then if I use the corresponding software. > > A pocket calculator is perfectly capable, if used correctly, and without > any intervals, > of balancing your checkbook correctly; and so you are by hand if you > take the time. > > > Calculator is capable of correct (guaranteed) computations, but we need to > > program and modify it accordingly, or at least explain. > > That is not necessary for just checkbook calculations. Better examples > are needed > to generate "demand" from prospective.customers. > > Ray Moore From owner-reliable_computing [at] interval [dot] louisiana.edu Sat Feb 5 14:35:55 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15KZtPs004605 for ; Sat, 5 Feb 2005 14:35:55 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j15KZtg1004604 for reliable_computing-outgoing; Sat, 5 Feb 2005 14:35:55 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15KZktX004600 for ; Sat, 5 Feb 2005 14:35:52 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j15KZgm10081; Sat, 5 Feb 2005 13:35:42 -0700 (MST) Message-Id: <200502052035.j15KZgm10081 [at] cs [dot] utep.edu> Date: Sat, 5 Feb 2005 13:35:41 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: Re: calculators To: vladik [at] cs [dot] utep.edu, reliable_computing [at] interval [dot] louisiana.edu, rmoore17 [at] columbus [dot] rr.com Cc: pjwilkinson [at] pasadena [dot] edu MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: ptRKLpr7lX1jsykyQ2swuQ== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk P.S. Same thing in the original example given by Dr. Wilkinson: if instead of 1.000001, students would see the interval [0.999999,1.00001] that clearly contain their own result 1, they would not perceive any discrepancy between what they did mentally and what the calculator did. On the other hand, with the existing calculator, when they get one result by thinking (1) and another result by using a calculator (1.000001), all they see is that there are two different results, so they are confused into thinking that one of these results is wrong. Of course, as Dr. Moore has absolutely correctly stated, the main point of Dr. Wilkinson's posting was that the students are unable to think about this problem at all. What I am saying is that even if they learn to think (but they are not 100% sure in themselves), and by thinking, they will come up with a correct result 1, they will still encounter potential confusion if they get a different number 1.000001 by using a calculator; this confusion will be elimintade (and thus, the students' progress improved) if the calculator, instead of providing the students with an approximate value with no guartantees of its accuracy, would istead provide them with a guiaranteed interval that is guaranteed to contain the actual result of the calculations. From owner-reliable_computing [at] interval [dot] louisiana.edu Sat Feb 5 16:07:02 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15M71nL004800 for ; Sat, 5 Feb 2005 16:07:01 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j15M71BM004799 for reliable_computing-outgoing; Sat, 5 Feb 2005 16:07:01 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j15M6qiF004795 for ; Sat, 5 Feb 2005 16:06:58 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j15M6g610517; Sat, 5 Feb 2005 15:06:42 -0700 (MST) Message-Id: <200502052206.j15M6g610517 [at] cs [dot] utep.edu> Date: Sat, 5 Feb 2005 15:06:41 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: Re: calculators To: vladik [at] cs [dot] utep.edu, reliable_computing [at] interval [dot] louisiana.edu, rmoore17 [at] columbus [dot] rr.com, pjwilkinson [at] pasadena [dot] edu MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: QWP5puicj43eCKgOlx6fUw== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear Dr. Williamson, I agree (and have always agreed) with your main point: that if people cannot think and cannot do the computations by hand, then they should learn how to do it. I am NOT arguing against it. I am all for it. It was never my intent to argue that students should not learn how to think or that they should not learn how to add by hand. They should, you are absolutely right. What I am making is another claim. But what I am doing is making one step forward. Let us consider the students who have learned how to think and how to do computations in their heads (and there are such students). For such students, I argue that interval calculator will be less confsuing than the existing one. This is what my argument was about. I hope that this (fifth?) time I have made my argument clearer. Vladik > User-Agent: Microsoft-Entourage/11.1.0.040913 > Date: Sat, 05 Feb 2005 13:25:51 -0800 > Subject: Re: calculators > From: "Paul Wilkinson, Ph. D." > To: Vladik Kreinovich , , > Mime-version: 1.0 > Content-transfer-encoding: 7bit > > Your example here of when a student sees that a calculator gives a different > answer than by doing the same task by hand, and then going on to rely on the > calculator's answer is how I have been able to get more change back from a > purchase than what it was supposed to be. For example, if the total was > $3.72 I have put down $4.02, and sometimes $4.22. The look on the people's > faces behind the counter when I do this is that of total confusion. They > then enter something into the "machine" and then give me what the "machine" > tells then to give me. On more than one occasion, even when pointing this > out to the store manager, I have been told that if the "machine" tells then > to give that amount of change, they have to give that change and that the > "machine is always right." One time, to see how far this perpetuated through > the organization, I called the district manager, and then the regional > manager. Both sent me complimentary food coupons for "my troubles" and > "assured" me that the "machine is always right." Finally, at a VP level, > things changed -- but they still sent me free coupons for "my troubles and > honesty." > > So, if even making change cannot be done correctly without a calculator, > which uses only two decimal digits and subtraction / addition, what good > would an interval calculator do to this people? > > (BTW The US Stock Exchange went to doing things in total decimal format > several years ago because it was "more accurate." I think the real reason > was that most new stock brokers could not do the fractions in their head, or > that most calculators of that time could not do fraction arithmetic. > Depending upon how many digits are used in the calculation, we are not > getting our full "share" of a stock purchase any more) > > > On 2/5/05 12:25 PM, "Vladik Kreinovich" wrote: > > > Dear Ray, Dear participants of the mailing list, > > > > I seem to be not in my clearest today, my apologies. > > > > Of course, I did not mean that interval calculators somehow help in balancing > > a > > checkbook. I apologize for creating this false impression. > > > > Let me try to state my case clearer. > > > > Let me describe a scenario how current calculators can cause confusion and how > > interval calculators can help. > > > > * First, a student balances a checkbook by hand and then again by using a > > calculator. In many such cases, the two results are somewhat different. > > > > So, the student checks again and again, and find the mistake in what he did by > > hand. > > > > This happens several times, with different examples. > > > > As a result, the student acquires the following rule: if a calculator produces > > one result and computations by hand produce a different result, then, most > > probably, a calculator is right and calculations by hand are wrong. > > > > * Now, a student computes the square root of 2 and then squares the resulting > > number. If he or she does it "by hand" (i.e., without using a calculator, in > > this case, by writing expressions on a sheet of paper, by using reasoning), > > this student should get the exact same value 2. > > > > The student is well aware that he sometimes makes mistakes. > > > > So, to be on the safe side, this students checks his or her result on the > > calculator. The result is not exactly 2, but something like 2.00000001. > > > > The student already knows, from earlier painful experience, that when the > > calculator and his manual results differ, there is a serious probability that > > the calculator is right. > > > > (Even when the results of, say, checkbook balancing were differing by 1 cent > > between what the student got by hand and what the student got by using a > > calculator, still the most usual case was a student's mistake, not the > > calculator's mistake.) > > > > This is where the student is confused: instead of trusting the results of his > > own work, he trusts the calculator instead, and this is a wrong occasion to > > trust the calculator: because with real numbers, calculators are only > > approximate devices. > > > > * How would an interval calculator help: to check the value 2 that the student > > got by arguing on a sheet of paper, the student runs the same computations on > > a > > calculator. As a result, the student gets an interval [1.99999999,2.000001] > > that clearly contains his result 2. > > > > So, instead of being confused (as with the traditional calculator), the > > student > > is reassured that his or her original sheet-of-paper calculations were > > correct. > > > > I hope I made my case clearer now. > > > > Vladik > > > > P.S. I agree that balancing a checkbook is not the right example for > > convincing > > the general audience, I am simply trying to explain why I forwarded the > > original email to the interval mailing list in the first place. > > > >> From: "Ray Moore" > >> To: "Vladik Kreinovich" , > > > >> Cc: > >> Subject: Re: calculators > >> Date: Sat, 5 Feb 2005 15:04:24 -0500 > >> MIME-Version: 1.0 > >> Content-Transfer-Encoding: 7bit > >> X-Priority: 3 > >> X-MSMail-Priority: Normal > >> X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 > >> X-Virus-Scanned: Symantec AntiVirus Scan Engine > >> > >> Vladik, > >> > >>> What I wanted to say that is that on top of this point, adding uncertainty > >>> to > >>> calculators - whether in explanations in the class or explicitly in > >>> hardware -- > >>> would further help students. > >> > >> HOW would it help them ? > >> > >>> In the example given by Dr. Wilkinson, yes, mental arithmetic would help, > >>> but > >>> if they try square root of 2 then they need to know that the result is > >>> also not > >>> exact. > >> > >> WHY would they "need to know that the result is also not exact" ? > >> > >>> From the viewpoint of a kid, a weight measured by a scale is more accurate > >>> that > >>> a weight guessed by a human being; a velocity measured by a radar is more > >>> accurate than estimated visually. A natural kid's conclusion is that > >>> whatever > >>> calcuator produces shouldbe more accurate than what we do by hand. This is > >>> one > >>> of the sources of confusion. > >> > >> HOW does this cause confusion ? > >> > >>> Honestly, when I balance by checkbook by hand, I trust myself less (and > >>> make > >>> mistakes more often) then if I use the corresponding software. > >> > >> A pocket calculator is perfectly capable, if used correctly, and without > >> any intervals, > >> of balancing your checkbook correctly; and so you are by hand if you > >> take the time. > >> > >>> Calculator is capable of correct (guaranteed) computations, but we need to > >>> program and modify it accordingly, or at least explain. > >> > >> That is not necessary for just checkbook calculations. Better examples > >> are needed > >> to generate "demand" from prospective.customers. > >> > >> Ray Moore > > > From owner-reliable_computing [at] interval [dot] louisiana.edu Sun Feb 6 04:40:39 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j16AedoS006298 for ; Sun, 6 Feb 2005 04:40:39 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j16Aec5M006297 for reliable_computing-outgoing; Sun, 6 Feb 2005 04:40:38 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j16AeavW006293 for ; Sun, 6 Feb 2005 04:40:36 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j16AeaOs006292 for reliable_computing [at] interval [dot] louisiana.edu; Sun, 6 Feb 2005 04:40:36 -0600 (CST) Received: from adminmail.paccd.cc.ca.us (adminmail.paccd.cc.ca.us [198.188.4.35]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j167pGR0005654 for ; Sun, 6 Feb 2005 01:51:22 -0600 (CST) Received: from Spooler by adminmail.paccd.cc.ca.us (Mercury/32 v4.01b) ID MO041F7A; 5 Feb 2005 23:44:07 -0800 Received: from spooler by adminmail.paccd.cc.ca.us (Mercury/32 v4.01b); 5 Feb 2005 23:43:48 -0800 Received: from [10.0.1.2] (64.170.175.75) by manage.paccd.cc.ca.us (Mercury 1.48); 5 Feb 05 23:43:29 -0700 User-Agent: Microsoft-Entourage/11.1.0.040913 Date: Sat, 05 Feb 2005 23:50:43 -0800 Subject: Re: calculators From: "Paul Wilkinson, Ph. D." To: Vladik Kreinovich , , Message-ID: In-Reply-To: <200502052206.j15M6g610517 [at] cs [dot] utep.edu> Mime-version: 1.0 Content-type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7bit Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk I do understand your point.. But, I think, where we differ is where calculators should be used. The longer I am in higher education, the more I am becoming convinced that calculators should be removed from all levels of education below the collegiate level, and then used only in "advanced" classes where they are used as tools to assist. Examples of such classes would be organic chemistry, mechanics, optics, etc. Just as a side note, one of the courses that I taught when I was teaching mathematics on a part-time basis was a preparation course for the CBEST (California Basic Education Skills Test) for elementary and secondary school teachers. In California all elementary and secondary teachers, before getting their "Clear Credential," must pass a state-administered test on basic skills (such as mathematics, writing, etc). Every individual that was in these classes already was a college graduate with the additional year of teacher preparation (and some had graduate degrees). All of them had great difficulty in doing the mathematics of the course without a calculator. The course content included addition, subtraction, multiplication, division, fractions, percentages (and conversion between fraction and percentage), and volume / area calculations. Several of the students had taken the CBEST at least once, and failed the mathematics portion. On 2/5/05 2:06 PM, "Vladik Kreinovich" wrote: > Dear Dr. Williamson, > > I agree (and have always agreed) with your main point: that if people cannot > think and cannot do the computations by hand, then they should learn how to do > it. I am NOT arguing against it. I am all for it. It was never my intent to > argue that students should not learn how to think or that they should not > learn > how to add by hand. They should, you are absolutely right. > > What I am making is another claim. > > But what I am doing is making one step forward. > > Let us consider the students who have learned how to think and how to do > computations in their heads (and there are such students). For such students, > I > argue that interval calculator will be less confsuing than the existing one. > > This is what my argument was about. > > I hope that this (fifth?) time I have made my argument clearer. > > Vladik > >> User-Agent: Microsoft-Entourage/11.1.0.040913 >> Date: Sat, 05 Feb 2005 13:25:51 -0800 >> Subject: Re: calculators >> From: "Paul Wilkinson, Ph. D." >> To: Vladik Kreinovich , > , >> Mime-version: 1.0 >> Content-transfer-encoding: 7bit >> >> Your example here of when a student sees that a calculator gives a different >> answer than by doing the same task by hand, and then going on to rely on the >> calculator's answer is how I have been able to get more change back from a >> purchase than what it was supposed to be. For example, if the total was >> $3.72 I have put down $4.02, and sometimes $4.22. The look on the people's >> faces behind the counter when I do this is that of total confusion. They >> then enter something into the "machine" and then give me what the "machine" >> tells then to give me. On more than one occasion, even when pointing this >> out to the store manager, I have been told that if the "machine" tells then >> to give that amount of change, they have to give that change and that the >> "machine is always right." One time, to see how far this perpetuated through >> the organization, I called the district manager, and then the regional >> manager. Both sent me complimentary food coupons for "my troubles" and >> "assured" me that the "machine is always right." Finally, at a VP level, >> things changed -- but they still sent me free coupons for "my troubles and >> honesty." >> >> So, if even making change cannot be done correctly without a calculator, >> which uses only two decimal digits and subtraction / addition, what good >> would an interval calculator do to this people? >> >> (BTW The US Stock Exchange went to doing things in total decimal format >> several years ago because it was "more accurate." I think the real reason >> was that most new stock brokers could not do the fractions in their head, or >> that most calculators of that time could not do fraction arithmetic. >> Depending upon how many digits are used in the calculation, we are not >> getting our full "share" of a stock purchase any more) >> >> >> On 2/5/05 12:25 PM, "Vladik Kreinovich" wrote: >> >>> Dear Ray, Dear participants of the mailing list, >>> >>> I seem to be not in my clearest today, my apologies. >>> >>> Of course, I did not mean that interval calculators somehow help in > balancing >>> a >>> checkbook. I apologize for creating this false impression. >>> >>> Let me try to state my case clearer. >>> >>> Let me describe a scenario how current calculators can cause confusion and > how >>> interval calculators can help. >>> >>> * First, a student balances a checkbook by hand and then again by using a >>> calculator. In many such cases, the two results are somewhat different. >>> >>> So, the student checks again and again, and find the mistake in what he did > by >>> hand. >>> >>> This happens several times, with different examples. >>> >>> As a result, the student acquires the following rule: if a calculator > produces >>> one result and computations by hand produce a different result, then, most >>> probably, a calculator is right and calculations by hand are wrong. >>> >>> * Now, a student computes the square root of 2 and then squares the > resulting >>> number. If he or she does it "by hand" (i.e., without using a calculator, > in >>> this case, by writing expressions on a sheet of paper, by using reasoning), >>> this student should get the exact same value 2. >>> >>> The student is well aware that he sometimes makes mistakes. >>> >>> So, to be on the safe side, this students checks his or her result on the >>> calculator. The result is not exactly 2, but something like 2.00000001. >>> >>> The student already knows, from earlier painful experience, that when the >>> calculator and his manual results differ, there is a serious probability > that >>> the calculator is right. >>> >>> (Even when the results of, say, checkbook balancing were differing by 1 > cent >>> between what the student got by hand and what the student got by using a >>> calculator, still the most usual case was a student's mistake, not the >>> calculator's mistake.) >>> >>> This is where the student is confused: instead of trusting the results of > his >>> own work, he trusts the calculator instead, and this is a wrong occasion to >>> trust the calculator: because with real numbers, calculators are only >>> approximate devices. >>> >>> * How would an interval calculator help: to check the value 2 that the > student >>> got by arguing on a sheet of paper, the student runs the same computations > on >>> a >>> calculator. As a result, the student gets an interval [1.99999999,2.000001] >>> that clearly contains his result 2. >>> >>> So, instead of being confused (as with the traditional calculator), the >>> student >>> is reassured that his or her original sheet-of-paper calculations were >>> correct. >>> >>> I hope I made my case clearer now. >>> >>> Vladik >>> >>> P.S. I agree that balancing a checkbook is not the right example for >>> convincing >>> the general audience, I am simply trying to explain why I forwarded the >>> original email to the interval mailing list in the first place. >>> >>>> From: "Ray Moore" >>>> To: "Vladik Kreinovich" , >>> >>>> Cc: >>>> Subject: Re: calculators >>>> Date: Sat, 5 Feb 2005 15:04:24 -0500 >>>> MIME-Version: 1.0 >>>> Content-Transfer-Encoding: 7bit >>>> X-Priority: 3 >>>> X-MSMail-Priority: Normal >>>> X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 >>>> X-Virus-Scanned: Symantec AntiVirus Scan Engine >>>> >>>> Vladik, >>>> >>>>> What I wanted to say that is that on top of this point, adding > uncertainty >>>>> to >>>>> calculators - whether in explanations in the class or explicitly in >>>>> hardware -- >>>>> would further help students. >>>> >>>> HOW would it help them ? >>>> >>>>> In the example given by Dr. Wilkinson, yes, mental arithmetic would help, >>>>> but >>>>> if they try square root of 2 then they need to know that the result is >>>>> also not >>>>> exact. >>>> >>>> WHY would they "need to know that the result is also not exact" ? >>>> >>>>> From the viewpoint of a kid, a weight measured by a scale is more > accurate >>>>> that >>>>> a weight guessed by a human being; a velocity measured by a radar is more >>>>> accurate than estimated visually. A natural kid's conclusion is that >>>>> whatever >>>>> calcuator produces shouldbe more accurate than what we do by hand. This > is >>>>> one >>>>> of the sources of confusion. >>>> >>>> HOW does this cause confusion ? >>>> >>>>> Honestly, when I balance by checkbook by hand, I trust myself less (and >>>>> make >>>>> mistakes more often) then if I use the corresponding software. >>>> >>>> A pocket calculator is perfectly capable, if used correctly, and > without >>>> any intervals, >>>> of balancing your checkbook correctly; and so you are by hand if you >>>> take the time. >>>> >>>>> Calculator is capable of correct (guaranteed) computations, but we need > to >>>>> program and modify it accordingly, or at least explain. >>>> >>>> That is not necessary for just checkbook calculations. Better examples >>>> are needed >>>> to generate "demand" from prospective.customers. >>>> >>>> Ray Moore >>> >> > From owner-reliable_computing [at] interval [dot] louisiana.edu Sun Feb 6 05:03:59 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j16B3xEx006429 for ; Sun, 6 Feb 2005 05:03:59 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j16B3whI006428 for reliable_computing-outgoing; Sun, 6 Feb 2005 05:03:58 -0600 (CST) Received: from lakermmtao01.cox.net (lakermmtao01.cox.net [68.230.240.38]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j16B3oLW006424 for ; Sun, 6 Feb 2005 05:03:55 -0600 (CST) Received: from Inspiron-8200 ([68.226.133.93]) by lakermmtao01.cox.net (InterMail vM.6.01.04.00 201-2131-117-20041022) with SMTP id <20050206110344.VEED11542.lakermmtao01.cox.net@Inspiron-8200>; Sun, 6 Feb 2005 06:03:44 -0500 Message-Id: <2.2.32.20050206110339.00a0a4b8 [at] pop [dot] louisiana.edu> X-Sender: rbk5287 [at] pop [dot] louisiana.edu X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Sun, 06 Feb 2005 05:03:39 -0600 To: Vladik Kreinovich , vladik [at] cs [dot] utep.edu, reliable_computing [at] interval [dot] louisiana.edu, rmoore17 [at] columbus [dot] rr.com From: "R. Baker Kearfott" Subject: Re: calculators Cc: pjwilkinson [at] pasadena [dot] edu Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk All, Yes, I see "sloppy thinking" frequently on students' papers in upper division undergraduate courses. For example, when a student is exhibiting a computation of the form (3+2) * 5 + 4, I might see the following on the paper: "5 = 25 = 29" I suspect this is at least partially a consequence of repeatedly pushing a button on a calculator that, although labelled "=" actually has the functionality of "accumulate partial result". In my undergraduate numerical analysis course (as many on this list probably also teach), we teach roundoff error in introductory sessions. I assigned some problems from a text in which students were to do some floating point computations on a moderately sensitive polynomial, using a 3-digit "toy" system and first rounding, then chopping. Carefully reading the students' papers, about 70% missed it. Several came up to me beforehand, totally befuddled. I saw liberal use of the "=" symbol, both as in my example above and as a substitute for "approximately equal." In 30 years of teaching this course, my impression is that this is the worst I have seen it. (However, several students DID get it. In fact, one did the assignment by producing an object-oriented class in Matlab that did decimal arithmetic to a prescribed number of places and with a prescribed rounding mode. However, amazingly, that student is apparently weak with other mathematical concepts.) Best regards, Baker At 01:35 PM 2/5/2005 -0700, Vladik Kreinovich wrote: >P.S. Same thing in the original example given by Dr. Wilkinson: if instead of >1.000001, students would see the interval [0.999999,1.00001] that clearly >contain their own result 1, they would not perceive any discrepancy between >what they did mentally and what the calculator did. > >On the other hand, with the existing calculator, when they get one result by >thinking (1) and another result by using a calculator (1.000001), all they see >is that there are two different results, so they are confused into thinking >that one of these results is wrong. > >Of course, as Dr. Moore has absolutely correctly stated, the main point of Dr. >Wilkinson's posting was that the students are unable to think about this >problem at all. > >What I am saying is that even if they learn to think (but they are not 100% >sure in themselves), and by thinking, they will come up with a correct result >1, they will still encounter potential confusion if they get a different >number 1.000001 by using a calculator; this confusion will be elimintade (and >thus, the students' progress improved) if the calculator, instead of providing >the students with an approximate value with no guartantees of its accuracy, >would istead provide them with a guiaranteed interval that is guaranteed to >contain the actual result of the calculations. > > > > --------------------------------------------------------------- R. Baker Kearfott, rbk [at] louisiana [dot] edu (337) 482-5346 (fax) (337) 482-5270 (work) (337) 993-1827 (home) URL: http://interval.louisiana.edu/kearfott.html Department of Mathematics, University of Louisiana at Lafayette (Room 217 Maxim D. Doucet Hall, 1403 Johnston Street) Box 4-1010, Lafayette, LA 70504-1010, USA --------------------------------------------------------------- From owner-reliable_computing [at] interval [dot] louisiana.edu Tue Feb 8 13:36:29 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j18JaT0x012156 for ; Tue, 8 Feb 2005 13:36:29 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j18JaTKD012155 for reliable_computing-outgoing; Tue, 8 Feb 2005 13:36:29 -0600 (CST) Received: from mx2.mail.ru (mx2.mail.ru [194.67.23.122]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j18JaJ9m012151 for ; Tue, 8 Feb 2005 13:36:25 -0600 (CST) Received: from [62.152.83.86] (port=1191 helo=slava) by mx2.mail.ru with smtp id 1Cyb9k-000PY8-00 for reliable_computing [at] interval [dot] louisiana.edu; Tue, 08 Feb 2005 22:36:17 +0300 Message-ID: <00b201c50e15$5f996e40$4652983e@slava> From: "Slava Nesterov" To: "RC mailing list" Subject: Reliable Computing, Vol.11, issue 3, 2005 Date: Tue, 8 Feb 2005 22:30:42 +0300 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_00A6_01C50E2D.D2C1F190" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1106 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Spam: Not detected Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk This is a multi-part message in MIME format. ------=_NextPart_000_00A6_01C50E2D.D2C1F190 Content-Type: text/plain; charset="koi8-r" Content-Transfer-Encoding: quoted-printable Reliable Computing Volume 11, issue 3, 2005 Mathematical Research Enclosing Solutions of Singular Interval Systems Iteratively Goetz Alefeld, Guenter Mayer 165-190 Fast Inclusion of Interval Matrix Multiplication=20 Takeshi Ogita, Shin'ichi Oishi 191-205 Exact Bounds on Finite Populations of Interval Data Scott Ferson, Lev Ginzburg, Vladik Kreinovich, Luc Longpre, Monica = Aviles 207-233 Tolerances in Geometric Constraint Problems Johannes Wallner, Hans-Peter Schroecker, Shi-Min Hu 235-251 ------=_NextPart_000_00A6_01C50E2D.D2C1F190 Content-Type: text/html; charset="koi8-r" Content-Transfer-Encoding: quoted-printable
          &nbs= p; =20 Reliable=20 Computing
          =  =20 Volume 11, issue 3, 2005
 
          &nbs= p;=20 Mathematical Research
 
Enclosing Solutions of Singular = Interval Systems=20 Iteratively
Goetz Alefeld, Guenter Mayer
165-190
 
Fast Inclusion of Interval Matrix = Multiplication=20
Takeshi Ogita, Shin'ichi Oishi
191-205
 
Exact Bounds on Finite Populations of = Interval=20 Data
Scott Ferson, Lev Ginzburg, Vladik Kreinovich, Luc Longpre, = Monica=20 Aviles
207-233
 
Tolerances in Geometric Constraint=20 Problems
Johannes Wallner, Hans-Peter Schroecker, Shi-Min=20 Hu
235-251
------=_NextPart_000_00A6_01C50E2D.D2C1F190-- From owner-reliable_computing [at] interval [dot] louisiana.edu Tue Feb 15 23:27:50 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1G5RoNx028120 for ; Tue, 15 Feb 2005 23:27:50 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1G5RoI9028119 for reliable_computing-outgoing; Tue, 15 Feb 2005 23:27:50 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1G5RfVh028114 for ; Tue, 15 Feb 2005 23:27:47 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j1G5RBt19348; Tue, 15 Feb 2005 22:27:11 -0700 (MST) Message-Id: <200502160527.j1G5RBt19348 [at] cs [dot] utep.edu> Date: Tue, 15 Feb 2005 22:27:11 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: a book of potential interest to interval researchers To: reliable_computing [at] interval [dot] louisiana.edu Cc: chleb [at] math [dot] cas.cz MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: eUVL2bxhxe4PovM/tDhigQ== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Forwarding, from NA Digest ****************************************** From: Jan Chleboun Date: Wed, 9 Feb 2005 12:09:50 +0100 (CET) Subject: New Book on Uncertain Input Data Uncertain Input Data Problems and the Worst Scenario Method by Ivan Hlavacek, Jan Chleboun, and Ivo Babuska; North-Holland Series in Applied Mathematics and Mechanics, Volume 46, Elsevier, Amsterdam, December 2004. Hardback, xxvi+458 pages, ISBN: 0-444-51435-X This book deals with the impact of uncertainty in input data on the outputs of mathematical models. Uncertain inputs as scalars, tensors, functions, or domain boundaries are considered. In practical terms, material parameters or constitutive laws, for instance, are uncertain, and quantities as local temperature, local mechanical stress, or local displacement are monitored. The goal of the worst scenario method is to extremize the quantity over the set of uncertain input data. A general mathematical scheme of the worst scenario method, including approximation by finite element methods, is presented, and then applied to various state problems modeled by differential equations or variational inequalities: nonlinear heat flow, Timoshenko beam vibration and buckling, plate buckling, contact problems in elasticity and thermoelasticity with and without friction, and various models of plastic deformation, to list some of the topics. Dozens of examples, figures, and tables are included. More details can be found at http://books.elsevier.com/elsevier/?isbn=044451435X Jan Chleboun Mathematical Institute Academy of Sciences Zitna 25 ( TeX: \v{Z}itn\'a ) 115 67 Prague 1 Czech Republic e-mail: chleb [at] math [dot] cas.cz From owner-reliable_computing [at] interval [dot] louisiana.edu Wed Feb 16 06:38:02 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GCc2J9029074 for ; Wed, 16 Feb 2005 06:38:02 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1GCc2EC029073 for reliable_computing-outgoing; Wed, 16 Feb 2005 06:38:02 -0600 (CST) Received: from ms-smtp-03-eri0.ohiordc.rr.com (ms-smtp-03-smtplb.ohiordc.rr.com [65.24.5.137]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GCbrdY029069 for ; Wed, 16 Feb 2005 06:37:59 -0600 (CST) Received: from Moore (wor226111.columbus.rr.com [204.210.226.111]) by ms-smtp-03-eri0.ohiordc.rr.com (8.12.10/8.12.7) with SMTP id j1GC0jwZ018590; Wed, 16 Feb 2005 07:00:45 -0500 (EST) Message-ID: <000701c5141f$364f8550$1702a8c0@Moore> From: "Ray Moore" To: "Vladik Kreinovich" , Cc: References: <200502160527.j1G5RBt19348 [at] cs [dot] utep.edu> Subject: Re: a book of potential interest to interval researchers Date: Wed, 16 Feb 2005 07:00:51 -0500 MIME-Version: 1.0 Content-Type: text/plain; format=flowed; charset="iso-8859-1"; reply-type=original Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.2180 X-MIMEOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 X-Virus-Scanned: Symantec AntiVirus Scan Engine Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk It seems odd to me that there is no mention of interval computation in the book. Ray Moore ----- Original Message ----- From: "Vladik Kreinovich" To: Cc: Sent: Wednesday, February 16, 2005 12:27 AM Subject: a book of potential interest to interval researchers > Forwarding, from NA Digest > ****************************************** > From: Jan Chleboun > Date: Wed, 9 Feb 2005 12:09:50 +0100 (CET) > Subject: New Book on Uncertain Input Data > > Uncertain Input Data Problems and the Worst Scenario Method > by Ivan Hlavacek, Jan Chleboun, and Ivo Babuska; > North-Holland Series in Applied Mathematics and Mechanics, Volume 46, > Elsevier, Amsterdam, December 2004. > Hardback, xxvi+458 pages, ISBN: 0-444-51435-X > > This book deals with the impact of uncertainty in input data on the > outputs of mathematical models. Uncertain inputs as scalars, tensors, > functions, or domain boundaries are considered. In practical terms, > material parameters or constitutive laws, for instance, are uncertain, and > quantities as local temperature, local mechanical stress, or local > displacement are monitored. The goal of the worst scenario method is to > extremize the quantity over the set of uncertain input data. > > A general mathematical scheme of the worst scenario method, including > approximation by finite element methods, is presented, and then applied to > various state problems modeled by differential equations or variational > inequalities: nonlinear heat flow, Timoshenko beam vibration and buckling, > plate buckling, contact problems in elasticity and thermoelasticity with > and without friction, and various models of plastic deformation, to list > some of the topics. Dozens of examples, figures, and tables are included. > > More details can be found at > http://books.elsevier.com/elsevier/?isbn=044451435X > > Jan Chleboun > Mathematical Institute > Academy of Sciences > Zitna 25 ( TeX: \v{Z}itn\'a ) > 115 67 Prague 1 > Czech Republic > e-mail: chleb [at] math [dot] cas.cz > > From owner-reliable_computing [at] interval [dot] louisiana.edu Wed Feb 16 09:26:51 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GFQpoe029452 for ; Wed, 16 Feb 2005 09:26:51 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1GFQo3g029451 for reliable_computing-outgoing; Wed, 16 Feb 2005 09:26:50 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GFQgju029447 for ; Wed, 16 Feb 2005 09:26:47 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j1GFQao21752; Wed, 16 Feb 2005 08:26:36 -0700 (MST) Message-Id: <200502161526.j1GFQao21752 [at] cs [dot] utep.edu> Date: Wed, 16 Feb 2005 08:26:36 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: Re: a book of potential interest to interval researchers To: vladik [at] cs [dot] utep.edu, reliable_computing [at] interval [dot] louisiana.edu, rmoore17 [at] columbus [dot] rr.com Cc: chleb [at] math [dot] cas.cz MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: XTf0GpWmKSCp5/BA/vzTOg== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear Ray, I agree with you 100%. There may be a mention of intervals inside, but it is unusul that these techniques are not mentioned in the general description - while the topic seems to be taylor-made for the interval techniques. That is why I forwarded this information to the interval mailing list: that this is an overview of the problems where interval will probably help and are not under-used. As you may remember from some of my previous email, I am currently starting to work together with Dr. Pavel Solin, one of the former co-workers of Dr. Babuska (he invited Babuska to El Paso last year). From my experience of working with Dr. Solin, he is enthusaistic about all possible techniques, he is just not that familiar with interval techniques and their capabilities. I think this is one more opportunity for us to teach our techniques to a new set of folks, and by collaborating with them convince them that interval techniques should definitely be a part of their toolbox. In my experience of working with engineers and applied mathematicians, sometimes the problem is that they do not appreciate th eneed for validated guaranteed worts-case estimates. This book is an example of an area where the problem is well-understood and since this problem is exactly what our interval techniques are good at solving, there is a great potential of using these techniques. This is one of the reasons why I am so enthusiastic about this book and potential for usese of interval techniques. Vladik > From: "Ray Moore" > To: "Vladik Kreinovich" , > Cc: > Subject: Re: a book of potential interest to interval researchers > Date: Wed, 16 Feb 2005 07:00:51 -0500 > MIME-Version: 1.0 > Content-Transfer-Encoding: 7bit > X-Priority: 3 > X-MSMail-Priority: Normal > X-MIMEOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 > X-Virus-Scanned: Symantec AntiVirus Scan Engine > > It seems odd to me that there is no mention of interval computation in the > book. > > Ray Moore > > ----- Original Message ----- > From: "Vladik Kreinovich" > To: > Cc: > Sent: Wednesday, February 16, 2005 12:27 AM > Subject: a book of potential interest to interval researchers > > > > Forwarding, from NA Digest > > ****************************************** > > From: Jan Chleboun > > Date: Wed, 9 Feb 2005 12:09:50 +0100 (CET) > > Subject: New Book on Uncertain Input Data > > > > Uncertain Input Data Problems and the Worst Scenario Method > > by Ivan Hlavacek, Jan Chleboun, and Ivo Babuska; > > North-Holland Series in Applied Mathematics and Mechanics, Volume 46, > > Elsevier, Amsterdam, December 2004. > > Hardback, xxvi+458 pages, ISBN: 0-444-51435-X > > > > This book deals with the impact of uncertainty in input data on the > > outputs of mathematical models. Uncertain inputs as scalars, tensors, > > functions, or domain boundaries are considered. In practical terms, > > material parameters or constitutive laws, for instance, are uncertain, and > > quantities as local temperature, local mechanical stress, or local > > displacement are monitored. The goal of the worst scenario method is to > > extremize the quantity over the set of uncertain input data. > > > > A general mathematical scheme of the worst scenario method, including > > approximation by finite element methods, is presented, and then applied to > > various state problems modeled by differential equations or variational > > inequalities: nonlinear heat flow, Timoshenko beam vibration and buckling, > > plate buckling, contact problems in elasticity and thermoelasticity with > > and without friction, and various models of plastic deformation, to list > > some of the topics. Dozens of examples, figures, and tables are included. > > > > More details can be found at > > http://books.elsevier.com/elsevier/?isbn=044451435X > > > > Jan Chleboun > > Mathematical Institute > > Academy of Sciences > > Zitna 25 ( TeX: \v{Z}itn\'a ) > > 115 67 Prague 1 > > Czech Republic > > e-mail: chleb [at] math [dot] cas.cz > > > > From owner-reliable_computing [at] interval [dot] louisiana.edu Wed Feb 16 10:24:41 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GGOfOW029595 for ; Wed, 16 Feb 2005 10:24:41 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1GGOfJb029594 for reliable_computing-outgoing; Wed, 16 Feb 2005 10:24:41 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GGOWSv029590 for ; Wed, 16 Feb 2005 10:24:38 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j1GGORf22170; Wed, 16 Feb 2005 09:24:27 -0700 (MST) Message-Id: <200502161624.j1GGORf22170 [at] cs [dot] utep.edu> Date: Wed, 16 Feb 2005 09:24:27 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: a course of potential interest to interval researchers To: reliable_computing [at] interval [dot] louisiana.edu Cc: mceberio [at] utep [dot] edu MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: +AhxrwbWrsBRJfWo9oDWOA== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear Friends, As many of you know, I have recently joined a group of people interested in extended interval techniques to the situations when, in addition to guaranteed (interval) bounds, we also have some partial information about the probability of different values within the corresponding intervals. This activivity was pioneered by Ray Moore himself in his 1980s papers, Dan Berleant has a website (easily acceissble from the interval webpage) on probabilities and intervals. For control applications, this activity was done in 1990s by Bob Barmish, the author of one of the most well-known books on interval-based (robust) control. >From this viewpoint, the following course may be of interest to many researchers from our community. Many thanks to Greg Coxson for brining this to our attention. ------------- Begin Forwarded Message ------------- Date: Wed, 16 Feb 2005 09:19:51 -0500 From: "Coxson, Gregory E" I wonder if those in the group would be interested in this that I took off the Control Theory eLetter. ... Course on Randomized Algorithms for Analysis and Control of Uncertain Systems Lecturer: Roberto Tempo IEIIT-CNR Politecnico di Torino, Italy tempo [at] polito [dot] it Date: March 14-18, 2005 Place : Dept. Ing. de Sistemas y Automatica Escuela Superior de Ingenieros, Univ. Sevilla, Spain Summary This course concentrates on nonstandard tools for control of uncertain systems with main emphasis on the interplay of probability and robustness. The objective is to combine hard bounds, which are frequently used in classical robust control, with probabilistic information which is often neglected in this context. The main advantage is to provide additional insight to the control engineer. This insight may be very useful in analyzing and designing complex control systems in the presence of uncertainty. The interplay of probability and robustness also leads to innovative concepts such as the probabilistic robustness margin and the probability degradation function. The algorithms obtained are low complexity (polynomial-time) and are associated to robustness bounds which are generally less conservative than the classical ones, obviously at the expense of a small risk expressed in probability. These algorithms are usually called "randomized algorithms." In the first part of the course, we concentrate on analysis and, in particular, we address the issue of finite sample size. Subsequently, we present results for sample generation in various norm-bounded sets of interest in robust control. These results are based on methods of statistical analysis and of the theory of random matrices. The construction of specific randomized algorithms concludes this part of the course. In the second part, we study probabilistic robust design of uncertain systems. We show how this problem can be formulated in the context of classical optimal control and then we discuss how randomization and stochastic gradient methods can be successfully used. We also consider extensions of this approach to linear parameter-varying systems. Other topics that will be addressed is the design of randomized algorithms for model predictive control and for robust fault-tolerant control. The course will end with a description of a number of open problems which may be important to consider in the near future. The course is focused on the exposition of the theoretical developments as well as on simulations showing the efficacy of these techniques. Main list of topics - Preliminaries and Motivations for a Probabilistic Approach - Uncertain Systems - The Interplay of Probability and Robustness - Randomized Algorithms - Sample Size Bounds and Statistical Learning Theory - Sample Generation Theory - Probabilistic Robust Design with Linear Quadratic Regulators - Probabilistic Design for Linear Parameter-Varying Systems - Randomized Algorithms for Model Predictive Control - Applications (robustness of high-speed networks, stability of quantized sampled-data systems, performance of flexible structures) - Discussion of Open Research Problems The course is funded by the Spanish Ministry of Education and and no fees will be charged. The number of places is limited and a first in first served system will be used. Registration: e-mail to eduardo [at] esi [dot] us.es ------------- End Forwarded Message ------------- From owner-reliable_computing [at] interval [dot] louisiana.edu Wed Feb 16 10:49:14 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GGnDse029756 for ; Wed, 16 Feb 2005 10:49:13 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1GGnDei029755 for reliable_computing-outgoing; Wed, 16 Feb 2005 10:49:13 -0600 (CST) Received: from mail-mta.sunlabs.com (dyn50.sunlabs.com [204.153.12.50]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GGn3ll029750 for ; Wed, 16 Feb 2005 10:49:09 -0600 (CST) Received: from mail.sunlabs.com ([152.70.2.186]) by mail-mta.sfvic.sunlabs.com (Sun Java System Messaging Server 6.1 HotFix 0.02 (built Aug 25 2004)) with ESMTP id <0IC00063UK15PV00@mail-mta.sfvic.sunlabs.com> for reliable_computing [at] interval [dot] louisiana.edu; Wed, 16 Feb 2005 08:48:41 -0800 (PST) Received: from [129.146.72.246] by mail.sunlabs.com (Sun Java System Messaging Server 6.1 HotFix 0.02 (built Aug 25 2004)) with ESMTPSA id <0IC000BK0K14MH10 [at] mail [dot] sunlabs.com> for reliable_computing [at] interval [dot] louisiana.edu; Wed, 16 Feb 2005 08:48:41 -0800 (PST) Date: Wed, 16 Feb 2005 08:48:44 -0800 From: "G. William Walster" Subject: Re: a book of potential interest to interval researchers In-reply-to: <000701c5141f$364f8550$1702a8c0@Moore> To: Ray Moore Cc: Vladik Kreinovich , reliable_computing [at] interval [dot] louisiana.edu, chleb [at] math [dot] cas.cz Reply-to: Bill.Walster [at] sun [dot] com Message-id: <4213796C.3080008 [at] sun [dot] com> Organization: Sun Microsystems Laboratories MIME-version: 1.0 Content-type: text/plain; charset=us-ascii; format=flowed Content-transfer-encoding: 7BIT X-Accept-Language: en-us, en References: <200502160527.j1G5RBt19348 [at] cs [dot] utep.edu> <000701c5141f$364f8550$1702a8c0@Moore> User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7) Gecko/20040616 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk I agree. I have ordered it from Amazon. Perhaps the authors don't know about intervals. Many people still do not. The encouraging thing is that even point people are starting to realize the importance of input uncertainty. This can only be good for us. :) Cheers, Bill P.S. If nothing else, this is a source of good examples for you to use in your talk. :) Ray Moore wrote: > It seems odd to me that there is no mention of interval computation in > the book. > > Ray Moore > > ----- Original Message ----- From: "Vladik Kreinovich" > To: > Cc: > Sent: Wednesday, February 16, 2005 12:27 AM > Subject: a book of potential interest to interval researchers > > >> Forwarding, from NA Digest >> ****************************************** >> From: Jan Chleboun >> Date: Wed, 9 Feb 2005 12:09:50 +0100 (CET) >> Subject: New Book on Uncertain Input Data >> >> Uncertain Input Data Problems and the Worst Scenario Method >> by Ivan Hlavacek, Jan Chleboun, and Ivo Babuska; >> North-Holland Series in Applied Mathematics and Mechanics, Volume 46, >> Elsevier, Amsterdam, December 2004. >> Hardback, xxvi+458 pages, ISBN: 0-444-51435-X >> >> This book deals with the impact of uncertainty in input data on the >> outputs of mathematical models. Uncertain inputs as scalars, tensors, >> functions, or domain boundaries are considered. In practical terms, >> material parameters or constitutive laws, for instance, are uncertain, >> and >> quantities as local temperature, local mechanical stress, or local >> displacement are monitored. The goal of the worst scenario method is to >> extremize the quantity over the set of uncertain input data. >> >> A general mathematical scheme of the worst scenario method, including >> approximation by finite element methods, is presented, and then >> applied to >> various state problems modeled by differential equations or variational >> inequalities: nonlinear heat flow, Timoshenko beam vibration and >> buckling, >> plate buckling, contact problems in elasticity and thermoelasticity with >> and without friction, and various models of plastic deformation, to list >> some of the topics. Dozens of examples, figures, and tables are included. >> >> More details can be found at >> http://books.elsevier.com/elsevier/?isbn=044451435X >> >> Jan Chleboun >> Mathematical Institute >> Academy of Sciences >> Zitna 25 ( TeX: \v{Z}itn\'a ) >> 115 67 Prague 1 >> Czech Republic >> e-mail: chleb [at] math [dot] cas.cz >> >> > From owner-reliable_computing [at] interval [dot] louisiana.edu Wed Feb 16 10:50:14 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GGoDR7029804 for ; Wed, 16 Feb 2005 10:50:13 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1GGoDQD029802 for reliable_computing-outgoing; Wed, 16 Feb 2005 10:50:13 -0600 (CST) Received: from mail-mta.sunlabs.com (dyn50.sunlabs.com [204.153.12.50]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GGo0KV029793 for ; Wed, 16 Feb 2005 10:50:06 -0600 (CST) Received: from mail.sunlabs.com ([152.70.2.186]) by mail-mta.sfvic.sunlabs.com (Sun Java System Messaging Server 6.1 HotFix 0.02 (built Aug 25 2004)) with ESMTP id <0IC000644K37PV00@mail-mta.sfvic.sunlabs.com> for reliable_computing [at] interval [dot] louisiana.edu; Wed, 16 Feb 2005 08:49:55 -0800 (PST) Received: from [129.146.72.246] by mail.sunlabs.com (Sun Java System Messaging Server 6.1 HotFix 0.02 (built Aug 25 2004)) with ESMTPSA id <0IC000BK6K36MH10 [at] mail [dot] sunlabs.com> for reliable_computing [at] interval [dot] louisiana.edu; Wed, 16 Feb 2005 08:49:55 -0800 (PST) Date: Wed, 16 Feb 2005 08:49:58 -0800 From: "G. William Walster" Subject: P.S. In-reply-to: <000701c5141f$364f8550$1702a8c0@Moore> To: Ray Moore Cc: Vladik Kreinovich , reliable_computing [at] interval [dot] louisiana.edu, chleb [at] math [dot] cas.cz Reply-to: Bill.Walster [at] sun [dot] com Message-id: <421379B6.6050504 [at] sun [dot] com> Organization: Sun Microsystems Laboratories MIME-version: 1.0 Content-type: text/plain; charset=us-ascii; format=flowed Content-transfer-encoding: 7BIT X-Accept-Language: en-us, en References: <200502160527.j1G5RBt19348 [at] cs [dot] utep.edu> <000701c5141f$364f8550$1702a8c0@Moore> User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7) Gecko/20040616 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk P.S. Did you have any comments about the short note I sent regarding the use of Popper's ideas? Cheers, Bill From owner-reliable_computing [at] interval [dot] louisiana.edu Wed Feb 16 11:04:39 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GH4dPx029984 for ; Wed, 16 Feb 2005 11:04:39 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1GH4cuV029983 for reliable_computing-outgoing; Wed, 16 Feb 2005 11:04:38 -0600 (CST) Received: from mail-mta.sunlabs.com (dyn50.sunlabs.com [204.153.12.50]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GH4INq029979 for ; Wed, 16 Feb 2005 11:04:28 -0600 (CST) Received: from mail.sunlabs.com ([152.70.2.186]) by mail-mta.sfvic.sunlabs.com (Sun Java System Messaging Server 6.1 HotFix 0.02 (built Aug 25 2004)) with ESMTP id <0IC00064NKQTPV00@mail-mta.sfvic.sunlabs.com> for reliable_computing [at] interval [dot] louisiana.edu; Wed, 16 Feb 2005 09:04:05 -0800 (PST) Received: from [129.146.72.246] by mail.sunlabs.com (Sun Java System Messaging Server 6.1 HotFix 0.02 (built Aug 25 2004)) with ESMTPSA id <0IC000BL9KQPMH10 [at] mail [dot] sunlabs.com> for reliable_computing [at] interval [dot] louisiana.edu; Wed, 16 Feb 2005 09:04:02 -0800 (PST) Date: Wed, 16 Feb 2005 09:04:05 -0800 From: "G. William Walster" Subject: Re: P.S. In-reply-to: <421379B6.6050504 [at] sun [dot] com> To: Bill.Walster [at] sun [dot] com Cc: Ray Moore , Vladik Kreinovich , reliable_computing [at] interval [dot] louisiana.edu, chleb [at] math [dot] cas.cz Reply-to: Bill.Walster [at] sun [dot] com Message-id: <42137D05.2040003 [at] sun [dot] com> Organization: Sun Microsystems Laboratories MIME-version: 1.0 Content-type: multipart/mixed; boundary="Boundary_(ID_U3KS6C0RmmCtPus/Fz0qAA)" X-Accept-Language: en-us, en References: <200502160527.j1G5RBt19348 [at] cs [dot] utep.edu> <000701c5141f$364f8550$1702a8c0@Moore> <421379B6.6050504 [at] sun [dot] com> User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7) Gecko/20040616 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk This is a multi-part message in MIME format. --Boundary_(ID_U3KS6C0RmmCtPus/Fz0qAA) Content-type: text/plain; charset=us-ascii; format=flowed Content-transfer-encoding: 7BIT Sorry, everybody, but Ray. I intended this only for him, for now. The note I referred to is still a "work in progress". Nevertheless, I would appreciate any feedback regarding it. Thanks in advance, Bill G. William Walster wrote: > > P.S. > > Did you have any comments about the short note I sent > regarding the use of Popper's ideas? > > Cheers, > > Bill > --Boundary_(ID_U3KS6C0RmmCtPus/Fz0qAA) Content-type: application/pdf; x-mac-type=0; x-mac-creator=0; name=draft03.pdf Content-transfer-encoding: base64 Content-disposition: inline; filename=draft03.pdf JVBERi0xLjMKMyAwIG9iaiA8PAovTGVuZ3RoIDI3OTcgICAgICAKL0ZpbHRlciAvRmxhdGVE ZWNvZGUKPj4Kc3RyZWFtCnjajVndj9vGEX/3X3GPJGCx3OV3XorEbuI0gOvaQoOi7gOPok5E KVEhqbtc//rOb2Z2Sd3JSBFEtzszOzvfM0ubu5j+M3cmqaK4zO5yk0VJVdw1xzc/bN/86cc0 uTNFZE12t90z6Xb3r+DH0JgyqMMkDvqQVlNok6ALTRHssX0OEyvbU5gkwUO4saUNZhAdAGgB yINtmHqAzYMhTEwwYvUMvMGBVPB80wxoFbzDqQGAY5jFwRn4C36YnDkRk4JkYvJPYRoLYAjT YOepGxC24b+3f70rbFSU5d3GmqgyJSv4OUyZVwH9CtGFhbjHWdaZlVAoq7kDIA7+ERqSw9mG z5YO59jmsFgqgjARyz5BGrE4SZKxxdOEZDN3m8RGsUlZtp9wIAoJFAe/hhaWzomL/I91Td4h 42ySuCICEycCAXoKS7oLwldkFCifRGVZke7k9bxQ51Y5E9zj1IifC35qvyX/JCYLDNYpXRC8 FWks1jGgMVaZ6lPclVGV51CHrikqS+okUVpkfN33ZJqU7tpwHG0MHInfkX5hx43JyVkbum/2 9vH8NsbYqEzpr43yivltiV9MYYNz8FFmKUawqZkh2Ma6JhtZxBLWF6Y5hXK/pcjCpVaxRxby zGIIJYQ0uEBEtWkFJ8udBJoY3YWbEuFPHE5gmqkodKAS5MnrZwUgBH24KZz0h5X0COrMwFoc SLhJfpW9SDVgk6h6cvB3/uUbBMzL//K9rZCLyOdXrEWtiKOF9lFJ/jMxBahLFdhbSHsw7SC7 3HrPRxmwoGYpEZvUCydG2AGUBY/Ib4GzFfjMzjEqr46Lc/a0oZRoWIRh5Vmx5eKveNEIN40r kCh/ccYvlX4CZSwX63pt+9dXsrSiz8MLFVObLj7P9OxaBu9iFzmejzPtMzshVvtLvPONPj5J tGEV5/Mq2im2UYQ3SZn9YagmVbyEqon/r1glVlSVstxwChqO2wyIpCw1MA+ra9dioqBQrV6I MpWZxc1VLtb0MTTpEvNJZV7l6Wu/r/L0m8oXa+XNlfLCenPD9rvVRY3LJU3T9DrAbqeGz/O0 eJkLKEHXuVD6XADuZfRaliNbotc4MVz0Yi30L/lefIlJrgymVwn6VmDatAx+YW+vq2sPhKHW W6ntxJrr0iKEX6211GRjkS5XusOLqjGsdFz4ZURHgFsp4ctCkhk9t4paHzxXTkukwnF7qai9 JNxeZEWUkSly166SmIJuAINg6qa5PWFOmAU+H8a23sm6QyzFlHOHVhakbhJmefDTpdspaB7c OQV8bGkWKIMn/Azjf0KweRCamq9U5j/zZj+Mx3oG0aCXbcNSbIwJp2RbauiSTaIqlV6ZVNJ8 T0M/PCD7siL43MJUQVuP7mSSlatLafO+fQSq7Yfzke84YTsL8mucxR9JlJ+3n9/T2gj00zgQ 7GGsj7Kn0TH4p43jdONWGZklhQcoUichmlns1rsWAov5p7me251U1oeh7rGiqNgLZMdCORlD SWMxIMhgbJiVRi6p5smiHjbHdj4IyU7RTx1E4dWhc2bBFqzkxqNGwkyjUUDCKf5Qz158neZI fImALPjQwSgwMWn9bjh9tXHGc+OuZW5Nq0bR0GU5k4TLhuUQxsFzXzetcw8hOeAI3gzHM3I2 YFMBM4pvL/08CWCPLSQejgJwkn0BYtgvQViPrVAskUAbJnsWO5pVbVw8NbdHxkVEJMrkabBV x/LofMb9LP/E0h7PM8wCewLdTfJMkGCg6VNeAEXwAfEO42GzGE8Ke9CeGn9H/g11gPHqeDKv jogO7ohpgn54xwRfOKxfq9qwd6giLd5BL6p3Kgh2U71f7TjtMXVOrZyE4srk2E2TpDMRNM5P HdtGYqEHjmKXo/vcLwitBJMwejq0o7vvsLq8YQra/nZx0TbdiFUkFVw9db3zk093erd0YucB wo1to9Al0MSYYHCEs5aDMFIlIUwTVD3XlDo+VynLZNL0xWoR6KNWn5GKZS/3TG+FzayqQq5J /krGeDl9rNXOy1JXIbNDNcPDqfMxCJSzgZRnArATa6ctlDyo5Qkgs3qw5DZ3KhaZCvWtKizx c6x9zCaFZa4vKvipa9jrheHkZhPPogAdcFH6afsOGC2+RDs9TxLLoNI0IpZiRiGhMsmpwSTe EvB4V993HFucPM9hmfA7r0SBpSZWM82uc9jUYevT7ipHxHnnUSvrhVVlKz+uDtPoWNAEQAWC mukF5TfJFvuiqsKa4+AMygykeAiu/Z16saaR4eGbgGISrLwpwJcSlqkGKSqM90Wmm/ZSDQie vGpFSziOlD+dt9hxGQOSCn1ZFlQ5vkgV5O32wOGZ0gO11oVPrp2Q1ALmVJOyiJVWFuXKYUlE 9UXEgx6j8uPSQjRPUjVagYLCHVru1X54RCStFZSwFOMhzl1zWKoRDT6Ra1FUGFsXvkS8/QGb 96EJvpnJNESlcaEFWUKAALl0VUIsjWHoKVi5WqRxucRDpx2cYI3IcwWV5oG/juOu9RQPJz3o u6QbskDIccJmd5xOiqDAJQcfZFe/1uA6uxdleRTgme26OtCw1tzq/IqqdaBbekqsNRKeq6Vg 7ITGdY/1sLlqrtL1+b6Vyu1qmpQ5h8/7lDtGkhlbx3E/9JsbeaAVhC1oy0z7Nr/g0FpP9XwZ UbwA+y2UR3bQUq4Op+/kSf3B1aDleB50ruHzRxNbcQhL0e/ue23iswvi/KWHAZIABofpur/7 pJ1W6EHW35jclrLLRi8tWh2r4ltduZ7VyrXnSrErPQ38abV4aX1yHXimYS63w0LnKCIP/kwP bXqi/+2k3K8yCYBukYkLwkqCG5nOQWNz8tnMVN39ZXa11dLYhnoD9NhyQ6jvnaX76/Jvi/Kq OTiyqw5ii2I1fBFXaQ3ykSe4NHP3yJVLtcH99SSUzlRLkPL40YmsUo9Si5mEj3nLMivpHAPr uh5ZFx9PKz70Mg04AcrgV7mKJ7KrOBtrDkTMWmQ4luRe524dwWR4psVj1z5p15Wkw/dMaQbE c+0+baRGOgGR7dsnoXpoT60rwDIDZktHHCY3udya5fYjz/lUsX7B4RGnqd5+4u/i57MLVOGt r3L+hI4jZ1+3+mEaZExzoRYXvrVQSZ2abl3F6P2uj+FCH7r8kQFbbpYX39oqaW2E59YWy9c8 Zn3kIZficdyw4VXYpvNjf3MzXUWO/jLpQMwvGPdQOqrRoO5cr8damdmvSgnPcaMfLZrurNPn d/4Dt/9iACnKqCrlQ8Qn9g7NheQdmXp4sKQ8MHI0W31sKGyUJ7n+C8v3oeVPORnKvb6J2zDD vM2Djz6U6zHM+PsFNnuNC6QtkbxVEnndiwJyOnot90ZvZ+lFeHpZsfnYESq2fS12XkUUnCr2 lntdyqNCxo9L9Z125gyWZAx+dhe8XCjf2/5ZTp1ZH0U2qhsQomfv3CKJBwT7FHwfa49VCF+E Is02JOtLZmxU4qtwASnX7QwBudj0ka39xKbXjyA8uGWFGLbUCY8/gKDcpPLxLYC3UImedUqi AzwT0l+ho1GCShJMQZ6duF4vpDozENPu+jZ+M3dS9QSxX7k8+qOA/PsldNHNsfjqi1daRSnN L+LNpS8XOvvVkgS24ETGHzzGPwHFrZboZNAWpO+A2LjC+eASCa0QiGmQk7N/SmHHfWMSCv/Z hG0NyLoDPPsvDFniOzB72qZRlhRXnl59QElRp7khG2qoMqqmUZrhn8OqqCzl3zc4U9/8Zfvm f/m0/dplbmRzdHJlYW0KZW5kb2JqCjIgMCBvYmogPDwKL1R5cGUgL1BhZ2UKL0NvbnRlbnRz IDMgMCBSCi9SZXNvdXJjZXMgMSAwIFIKL01lZGlhQm94IFswIDAgNjEyIDc5Ml0KL1BhcmVu dCAxMyAwIFIKPj4gZW5kb2JqCjEgMCBvYmogPDwKL0ZvbnQgPDwgL0Y0MyA2IDAgUiAvRjQ3 IDkgMCBSIC9GNTMgMTIgMCBSID4+Ci9Qcm9jU2V0IFsgL1BERiAvVGV4dCBdCj4+IGVuZG9i agoxNiAwIG9iaiA8PAovTGVuZ3RoIDMwNjEgICAgICAKL0ZpbHRlciAvRmxhdGVEZWNvZGUK Pj4Kc3RyZWFtCnjarVlJj9s4Fr7nV9RRBmK3RO2dU6aRoDPAAAFSt8kcaJm2hdbi1lLu+vfz NlKU7Urn0CigTHF55Nu+9/gYPYXwFz1FcbzLs+IpS/OditOnqn33r+d3v3xO8qdyV2bx0/OR Zj4f/ht8HfrLZhuroB/rzTYKprrv4LuIg3jzv+d///I5jZdFudplcSYrP8K0tAims9mkUdAP myh45a56xN88+K5UvEnyYL9JVWCmCekbmDbAQAIfMKU+8tQauqf16hZI4jruNdjWY93IHkfd jPV3FaabKMoCOsJhxydOvBNvQQRZqZ62KtolaUIH/6RHPAkSyqPgiG3dwP7UiwdhsioJ9L7G gXraqILnKz4e/Hb9xA2N8srjoMdJf6EUWhIGyjEKdpttmoUkLJiLwtp6wkJ6vDNSMHqUzqnn X+LyaGce5Vc2PhnaYYtcP22jZFcmCfxGuzJNiVEz6MkQ7TJoiU3Yq9uonPcug8tgDpbpirSP 3PYdrilCOK2eeGJFTELXfrNVBekEPgzqw3JSBrQX0DAHZLpQwedNgWIBhfME4h0lpNsLNnEt 9b2H8TJk6dDGMHAma+kHIl4E/ZHYDFcMnlCOGhl68dWUxAmxVjueRuxMhSEcvuJUkM7An6Dx fKPiEJWYB9eRe0k2pNZrd7ZyahrgLger/nJkoixUapq/LlY+yDQsaA3JW3athXDFhGkd2O2M /6sJdYFUahrwuc2I2+sZGAJicJREhZYFK0PkJCqtOeGEfk/6MMMLziTqimXbc9PnGnjkzvnC rKJyGlBMkkSePjohgpqSBqkIt+6P3Ge1UgQvi0rWRztaa2xGawBJEqJf3CtZfGZgRpOkZDGm IRCAz0LYUCXveZU5Hf/ySWGy082A52iJZocLJp7YD7LA8Sr9ixTTjOVNEMnEtaz27Qk+byQr 553J6s8s2t0DVv/TD+QBOfq241o37GTgTOQoJSPVfEGGSrKjUQ6MSzX+ZA5q0P1eueve3Uc7 1+pjYhriiehXyDTuP3tboKa9rXgS2bveN4ZROIoWFI4jtSvCXALHX/cwXe7SJJXhFkVGu40T Q1us1C5O1UpUCxDFWdAw8oxbUEjHXahO/DV/zrrZEp4SANCKh3zhoiP/av75W74gTlCTD45S zoJf77mDY8epZe/L8Z4ONIrsR9IJ41KGKfzgUfejMxD2b4ISHrTsroR0GgwZKIlUxdkuhj3W jqa7BzyC6qLUsXh3uHSnEiXD79EgyzVcqDJisWPjVMNZJRDYsTXWK4gDAhQ4ehOdjewgQQVn jH1rpnPdnfjTW3sdeu4FrKzRxFdHuQwgJEBTcgTDGMwYwwIqdmGa38oH3SOGtXZyTwIWADck fWWNTAnIaHv4ebAzbq1v5O7+gWlEWb5LnG88MI5iV5bWdnR3eEChBE2XP6fAHcJXHHzp+EBe nMIgRmA02pj6nudQkpemwZ+WG0gKthQWajNKoheFojEWbRiBQatVbGv1hLbJqVAZsyfCL+uf ssMz92gJBNSuWHuXyVk/9i5kpmGWqYQHJWdUdnUnC4iI7jijg+/FbfAES6KpUoqi2Mv4RFZn t+weALrhRKcmDKlA4yCyGBOlyeaqq5Tcz1LfTsyTn0jMIwz0GGxIWJSfv3I3AQj8euk1fNEc 0BxGJpscR+h+5H3IK3Jwk28LQVRVWHJaCt8VkeqYwN7bYx4piT/wCGEU9B6Mo9rJxFZ3r2wq wpcvUgQ+Aj2CZL4bvOCtopv4nHR5EF+FQe3fImqbc1XYCYluPXLQ44tJx7ZNFK42UpLh4drp wf3CZt2kOD7gM7l/unjGKhKLR6jSHTGicJMKOBYEmri86tvLzAn1wBPG13EyLY9a+KKAPlcO RmCIHKSprXyUzVBZY6kPP/XBW/Ljy8Q425SdsgZETlKyyjx/wQMRHTw89fYjuQwM1Jxt9bwG dGwvIuBWDmhayWk6lzfj5HHe0+F8SQK6IALhxrpCymfehpAD1lzPS6rDSTP2Vss16N5bHR8q hVy99VAec1mbRKsUAyuntTix1X+gJCnUwKecscWvmK9ddF4ha7S7ZjavPOf2BurABjWIiBwV wdcNyZJRhxSLxH7/7ds3brF29lsLzjai4snALu55rbuqoTUHIxlLlLxh1oxH38Mof5SaLPnP R0Y3Vj0YJdgtWLsC6RfklnDxOevhwKkyJbpxTEAheUscyL1cBOGuNq2M9kfupiyfiEgoIMp8 DeTsAnsYzmHdj6J1wrkewvxi3x/uBbLl1haCaf7T8uiERTrApXFp93KJoAPgnidW1XRuudNW BaC5Z4ux55Vro3d0cWdcInP4nkE2Onphh+5q/xRzTPMmofsJdivRgrGfDAu67karN6ZjhoEc Z/ywtpR/TDcqC53wLdCDsfk2i1PqTsyZPw960uhzpTM/7JVUb+QcD3sk5YSW6XqyrZN8ayuY ah509crEEBdx0EP9B147symQH8yN6NXhRWc1PZg1WmMZinBdJOVFq0/awWOh5GIXexBCyRL0 WIDxK1g1RRwYlPKQ8stDsYt/teD6HZiSPiUgkqCZyFc8EKqeDUlu7aKWnNwCZr3UFh6u3IHf 9vhn88qdiDNvRbRWoxGqJMJciFphIGp85S9SSrIqV9RHxqgioWQ5iiHe43lnjEaoSOOqi7Ry lT1iPZErnRI+YG+SHTaW6APrOL+0WS32IFmcpvnz96/Pv3EHB9yxPr0R1U49GTmcNoPTrmpy JZZFtFNPg9erJIwJWGAMl7HXnWZ9MtxpkakMnYLl5BMDPmlTKrs4n0oK5SLDysW/elrKV0JS Cp4uTzvxYr5S2RKhK6A+KhZRUpWAd16YpwWGNi6JTsMFhkb+drEES0WCPEzo6FWf4B9qHe4R 34iLlvigtY0r0KQsQFhKOczcmkFAGLqWHG4RZXonShv+rEQfxG8pVZB44arP4q2kXr2WK0yz cqULS5zZO2p2V5HGLuv8Wmi7ergATIoAwxA02TiVk3wHoerlAAcevbpUGusrTWPrMZx63/Mn V7uZ7pyVeFSS28weW1QGBJy2iSZnT5QcYbfYEDYnV9CDZZeVPkcuoleyiGqY8Ovl4kSrE1De 3heOcLwFr59XdG3hd/ugkAv2hUA4jIzKD6BpQWVVxl4y3UlVhVJjfKhYMBEnSj6IF18vAXzl LlcZWZD5xE7g3STYf3O+1dEqDHslP3vgtwdvJb9T4O+DR5QFJ9Enwmip2cR2XJWJffjB3rNp LmyybyE2bpeE0Q+2o+EejHrg5nKxWGKJu1z4kT4JFYoPZ33jpdNNDs2PORn48DNln3IWAND2 QgVIJAHx5ld8/IicjD/aU8K1dardu9D4ll/DaBqGDieoTUwc/cQXkTpNbGHJGgtORv3QL/+g 7dSNGT7wk4xLt3GMfB8brba3Fj2t8LK5P4xEpsMj8O25fOR5P3cwDrJbcCnE2wAn3OIJiTrP k+DjKNV3AW8Uc5I4tMY2B9MLQSxfqpBijUrxq/w2luKgFjKXoXbFrQd+OgKOrmhQCFakI/yF zBw5oWbDrkQ30T9kHoIJ/nopL34ygimrvXM/Nwce4Zo/I89wlxUosDVTLbAa2Qzki+yj/+ZY j8xtZCD24Dnl4g3+3gife2vJ091rHl/456UGh8DuEkgOCelyo0n53pJSCYMYxMyjfXAOCVLp rRfXjffoZW9LnOea5vie1+D1755flnAl7wVc2+cnZfvokRZ2OweKhsf3VJowHENxGrsWvXZG efDMmdXyOn10dD0it0UhAmjJmlxA1Mtr+HIocSrKOMzyUKZ2cZk5XSYh3f7wnMORiLd2kc2n cErVDy5SVZP/vJQorhdhhMaZEm5Dp0BsPpDiK0+vZQYjdbK8nnnPFDiOJoa/xoZZeSSFJd/D NATHlGrRPMF3tOJ2lelBCCFzLR+aK/Se9cgNr9xfzYNzoAw6v8jcn7h5wH4vtblSolxyopw+ SpQTr7JRps5Csf2WJj2vvAUv7SIHleW1NLz3HXV/zVrKIeKW6GxSc4O2PhxuTYO6x+VhSRCX HjW4fmI3mkZvU77vG5vy+Pd95b1mR1m5i7MMrp/lrigK4lThyLtPz+/+D42SmlxlbmRzdHJl YW0KZW5kb2JqCjE1IDAgb2JqIDw8Ci9UeXBlIC9QYWdlCi9Db250ZW50cyAxNiAwIFIKL1Jl c291cmNlcyAxNCAwIFIKL01lZGlhQm94IFswIDAgNjEyIDc5Ml0KL1BhcmVudCAxMyAwIFIK Pj4gZW5kb2JqCjE0IDAgb2JqIDw8Ci9Gb250IDw8IC9GNDcgOSAwIFIgL0Y1MyAxMiAwIFIg L0Y0MyA2IDAgUiAvRjExIDE5IDAgUiAvRjE0IDIyIDAgUiA+PgovUHJvY1NldCBbIC9QREYg L1RleHQgXQo+PiBlbmRvYmoKMjUgMCBvYmogPDwKL0xlbmd0aCAzMTM5ICAgICAgCi9GaWx0 ZXIgL0ZsYXRlRGVjb2RlCj4+CnN0cmVhbQp42pVayY7cRhK96yv6YoAFqGjui405jD0SpAHG Y6D7ZvmQzcrqIsylxEWt/vuJLZlJFlvyoA/F3DMiI1+8iOzwLoC/8C6MYz/Pirsszf0oTu+q 9s0vD29+fJ/Ed6VfZvHdw5l6Ppz+8NThGKeRV/Xt9XAMvfoQeo0e/MMxDQLv4aKxOfbafoAG zX3ViD1HPUzYu++oxE3TRU38dVZ1g1+hN/WyhJp5HBf7R5pGD18Oaeapx0bq9ZdDVHi6O0S5 N41vsRI+YKELdedOsB99+PPh3yDEMQz9Mk1JmIo38ykK0kMYpd5JY7GrcFScwA46/ng8HHEN LlwbVekTf9fUIQY5zJCtXnDRMPHLJFmtTJpKQlJE6KlWStfLMrgfeyiBUC/YFHgVdey437Ih LKgrLdnAqBo3hnWoRPo167BoY/0kM6hOetbttdEtC46anNRU9wcsQDOoHT7PJMVxRwyWO088 heddTxeeCNVfV9wyvoyTbvE79uRAT9xCBw21VyNM6A20GnzgsoOMalWFerjUnSxmlDzDXnGl viMtohFmiff+UCTe0Mvg6VKPPEqWGY47hjBiY+npaqrZnMCO0jhAy8gPEbTgfkAx3dORp+lr NjhQUhmA9PVica2eUPSkDL2rEanHOb/UJz1iQ+TJ5eAzU7RmaSteeDTpQT9zQY1i/VZgnEix IfTzJN1WsmUk23Xon3BB1X6Kogh2ndPtS3LQI40+abo1gQeXFuqz9e2B8qBp8bmZRq6oJ/51 BKSJ6HwrPcJJJHAI71GaXtG1p4u20l7hTbwN0sZmBTXI2k9smXpQOFnzsnN4Yv+Lor/WrWMX cQnrnfl3pl38ZbZDvZ87blNVNZtpVPVyKGI0qCzOvH/KLAh9ZW5AY9SfZwsXofcW70ts4YgX Hk564GGDbpQA4LgnRNt3T6CcPFrbewQ2vJjLFU7CvakEgjjE2ilLDnZK/UXbJDCbK/am0SRs i1+P6/s38qKdppqTGYL6kvEjL0JHdn7hOkIC2j3h1AGhB7/2bpuG86krhXgPwGYPbwMEU41W EQNG0fmFoVUEzt9w3b4wA9lgnHv32hgyLPUfPnXYp9eTgcH4h1/+5XPrA0AF3oQwZVxH+JwH NhYskGRadbIp8l07wrEB8CnByScCkmUKljMsElQTN6I14cmQAhquY3NN7UA+XsAjtliuZGSm T8Xj/r4VYG+LC7I71XUGeieusn4P+rMb30XP5eKwyYA+C7jXHxGUsoLc2SifvAG99B0RU7Et Z5cFfbp+kgoyILWUajOJ/Ipc41jTPcAqe/uSrPTOCGg4VnGrVaPohzDpgELtqulbWAo6CRKL lERjghhUUQkGknRnRjVBdexAUn9drGCeyMZhqmfwnoTj2GtQojQmA7JRalKy0KYN6QZP9PHd O6x6d8zTRNb8yhB1NZ799ggNQqpOKAdC1l/L/CdCN+Q7VTOfbrZVFmufUaKuo/iQpraf/qqq CWphT2GeWj2QvjRurSO8LYRFwhyDVmMvSFoLCAtjxJruVH/B6Wk/M8DJrVjuie85b+RurwEI EzpEGoIK6MkGU6mrsM9Yrio0DXqBMT56IaSOmfEUK3WYaa8zsyYqMr9tZi2k5hUGSTSHDBzW WbSpmBk16NKTlBdPIvJd2G+xGsKyK12bqLAEuDPgypLWiNE4zmVchsTjwJ37zA10SSaZvlu2 J62f4ZNEJtqIw18jmG39dBGqELmAhAUkw/gL+uO9cVH0Z5bQ5AhC8OJMMVoZXVP3hAwLy6MR /SxWjJWISdjpuga4upK1VvefZluOXa+GjNI+oFJA2B/fh6ENruIg9+M4kxBr4A5u9JX7eVBI 83PdsPdzFUL6A8/lB6TAyIcbwtjcMyV/vmgy47wwNoifwt+LvX0bPd7sNgwz+CzXuw0T26Hw gUBIc8TNhTM88MPENP+R3M4PwiZm9p9vhic+WJC0Zn/CZrPM+2FnkzALlLnfT7cazfx0mea+ 32hjHkULUMqdq+xypNWA5UOUpm+ljgu/CI1Y6Q+3O4Id4+FJmA3hGZ9plPtZZs40s1FXVIR0 LWcTcGHchjeiCNBNkV1UwuFGrmYXI1AT5eVrIJxEmfex4yXY7eLosW6XOFPWBxDD6PEt2mKA GEOLGyXRSie9caYoRk5izBLSIJ2K3ZuOxW8GhWEs5EBzYQ+e4hDdM9KlehGQe6MadiZxcPfG mqI09ct8Y/KFa7JpbFr/gbotvGTniCO4ONH6hDdWW/phlv8/y2Q7th/FfvG65UNYFTg4yyb1 QZhlVLgEFjWPjDhIjR2IJ8ZfoG4LECFr49+nWQ2KwxNNIFpuQTK3ao+IOJiFZQZr1JgoiIgH MMPE5SZZnQJE/DARMBFusvSSw/U9miPpEAhh9MTpD3TglA9pGknBWFZZN/UShuCguWuE162S K9H6HpqwCXssO584iVPLSMU/bW8Zd99R2iAUW+dQtdoyNnt7+vMeMvv5YmA7bqRw7IdzEnTU wwrESs9SYJNta25NMfWLKP0Oihd/F8VvwTD00zB3rgptrBOg3sOtkbtclJWLOz9yaodCcuhA zKQel0glhbqPZ26zJNb4SjXsuOsIoLoojJ67vSsWg/hxuIY7PlE9TMp44DiyhNDhgIHFIrI2 vIKxmJsMXEomzxqHdq+u28lyuOv52hmHAbi/Mk/4IEqQJPazMHtNmsLPIxBp6bLcEEMmH5ly 8i7sXXCTMHHkZkZws3yR5tbQY2aARehn4Pxvc4zAtS2ZmhtDnUf+VfzTmttk0hUm3SwuAGc5 M1JUS56ylySviaeR9Dd7tCfP/DD4xqmXvmE1DiN9jdFDEBT4SVqsREWni0kvNymgxgnPPym8 1gZnSbqXEaShSuawsTum0+apN0kXyqLRFE+aQ5FBubhF/flUMZVAQiTpdxjiLnm30TLCXppL 0jl1PECSbtKLWPFKWnE9drGG89C3mHEIEycmglksOUmK3KvPMtp1N1hxqj9FSXIoI4ex78T/ Eo4osEab4eIkVpJnFpE2YeAouVqkkV9rio6gt2UzXCZjdH0IJ0XyXFwMpmz5jSDP19cA89Qh Zd86bpXM6u4ZrWVxrTIOxO0MPQeHWFGZXRpyRbvCFvG5RKMCpIGYmFzgwLzw0AVJ4/UNihbc bOcDxUW3LAmimMJeJZZnGCVtRuFczJAdLDBI2ZXAO/Vc2/V0aQKTNsK6/pFzV8coCP04S1eU etwo7HHFdDevTEx+z6oZ6yVy1iN3XLi5mCk9hL2YKjX9hD4e/MqnIA2l1smccNklREW4Z1uk YLEtXFUNdquOQQ+rDdOc5B5fDnI+wilKgFyC2n4b7muuYDOEkWKGUWHN52c0vETudmQyNDjq kYXkCcyFx+l3DWvJT7yY1NcCNi23XtBp4BImkbqAuPMuYM/05mWLHp5SsRn82CM69BJCWZ+b Nx1seW2HaJYQj/+G9j/J9OvsIAxeKiS7h53w4XMe9HHqj9u7a0z+e1lm85RSbm3SvLQm5eZN p0Qd9MML4uVbfrJ13DM0I1zSr+PJoLjNrcjk8lpTvqLQZMvSraspd19v9pLMy6skiIS0DQz6 v0Lbqn4WeEBZIBI1/KrYUTnUVWoUpstUe3B434YBs1Ig1HlSlixgB6Gbg/4814PrPZFGCk2t x76hJBhWPl/AhGCyy46l8n1DBXXr3CFcoIAMjW9SnJacccwCMWL+wJOkxnrEEYwGMgR6oFVg GiyNYu+3TXgn+NlYVEEdZpmdn/3dlWDp5XiDxxbxjyvJXrdT5m033iHOFv50HQ744n9IwyXM 2YbRSxwjCWGYlFNd9ESvBqkz+dlpZQNxYrUmSVmpJlU5LC0p/KhMV6Eyk7Tc+x0PqTewbF6X 12+smfel1s9vN7df8wQ2C0pP96xgHkXbhi5angLNe+0SGctLrXlcx/9hoHUL87b74df7e/6S EyZHxC9qUCmxwi5tk7yMvF1w4v/E113+fQCoYw+K4s+JiRp+mqfR+kT/EpIVpffA3gDfDtCy kji9mcsKiaUPvz/8yl8snKtQWmNxHs6jzR5gdI7VkVEH4qQ2KYh80SJ+K8ogZJTMguLzpW+k 17PxB1hobaKcKTmO6R/lH12cf1HBzut/UZENrJMiQtyxxclx5/Y/JnYAsWfrJcw/iynDseOI e+ffUuzRy1tCBoFKlsFcQeFnAXv+GJvevHt48z83/AWsZW5kc3RyZWFtCmVuZG9iagoyNCAw IG9iaiA8PAovVHlwZSAvUGFnZQovQ29udGVudHMgMjUgMCBSCi9SZXNvdXJjZXMgMjMgMCBS Ci9NZWRpYUJveCBbMCAwIDYxMiA3OTJdCi9QYXJlbnQgMTMgMCBSCj4+IGVuZG9iagoyMyAw IG9iaiA8PAovRm9udCA8PCAvRjQzIDYgMCBSIC9GMTEgMTkgMCBSIC9GMTQgMjIgMCBSIC9G OCAyOCAwIFIgL0YxMCAzMSAwIFIgL0Y1MyAxMiAwIFIgPj4KL1Byb2NTZXQgWyAvUERGIC9U ZXh0IF0KPj4gZW5kb2JqCjMwIDAgb2JqIDw8Ci9MZW5ndGgxIDcxMzIKL0xlbmd0aCAzNDkw ICAgICAgCi9GaWx0ZXIgL0ZsYXRlRGVjb2RlCj4+CnN0cmVhbQp42u1YfWgcxxV/e/txkizJ +rAtx5Lrkc6SWmvv9GFbsvwp3VmybMe2bKm25DiWVnd7dxvd7V5293S+hNLQYEycmjhNCcG4 IZgQTAjlbELSmhCahJCWtJS2lOYDWkoKTcCUEEho6kTqm9m9D0l2k/5RKKV7ut03M2/e+73f ezM7OuAAoBoeAh6mj4x19tQHfvAXAOkd7D0+Grp7/NJPEx+ifBO/veGkkvqs45NhgLXPAPDb w3M2eSr6w78CeL4A4FqiqVjSU/u6gvJG/L4TS2Sjf+td+x6AMII634urSiRujNwC8HZRe3Hs qP6S+xG249jeGE/aZ2ouc+9j+/vYrkkYYYWLwh8Qzy+xXZFUzqS4LVw9jl/GNtGVpOr35Oj8 V1D99ZRh2Quf4gisfYiOp0w19fKloTPon2C7DTBWOuB5nj9HtUSMIpxMan1QvBbw6ltYmHda t24N4H1gzepV9XW1NSurqypXVJSXeSVR4D0cyNfRXIicnznvC0aP++XrmwBC42dy0DHR0D/J HjvZI1exaTLHhWqWdPnlDvzLeTb5ZefGzeTAF7zG8a1BMkziSiTHt+KT5ISQTyGRawMD58cn mvG6Bs1Nk6MTVBBayXA8520dUnJwdKK5MSeGtk02KpHJsQlfp+/zxk7fR+i7487WlcgdNcXW 4J6Plk3e8znrqq1r6PfLr2HYyGuIUJDnh3J8SNGm904H/fIQyfG+4HXOI4bmQr7QeR8O5ARf cG1uBY0Vra0d0nIDCmG4c4ONKFMxN9yYG5mczIm+IJrgfEEt5/EFc5Wb1vplMhTf65dvQC2I Cz8LIojrtSLQJyHXPPz5oC/HKcFhJGxgfCI3MD1J/fllHArVhhBuN/jlgF9+0y/vya3a5AcJ yrD2W6ALlPnT89YCVhp/lj8nxsVnuPVcNVb97+EX8Brk4AV4Dp6FK/A0PAUX4FF4BNJgQBIi MAPTcApG4TAcgBHYC0HYA7ugH/qguexG2Tnv+95hb7O3RjonZaRTUlDySxukRqlaksS/i++K vxZviI+IEbFP+EB4Q3hJuCJcFmaEk8IBISjsEO4SGgTgb/J/5t/iX+Jf5F/gM3yS7/Tc54lw j3GnuB2I8E/wLq6Q38Fv4FfwNrwFb8IbiPhVeAleRNw/RuTPw1UX/WW4hPifhCfgcXgMHobv woOQhQzch3EcK8Qw7EbRD81Y+02SX2z3ttQ0Cf5sdVXVo1XZcmv3rJDNdlWF4MJP+J1n13d0 0EVFNxIRP7iTeKFloLHTI4EkdEI5eD3CRR6HLno5kDzlK/mObujoaq5trm3HLwcCNH8RFOEW NAuv3gqCBx5cuCk8xp+FLXAIegc6v7mhaaWwa1W5xFVwuwJ8hVjWyh3kOrhV/pbydWh824a6 KqEuPdwFIncErePVxe3mNvesaehjj9WrvLy0etXmnt1cX+8OzxrMLedraWuv5ryO0LZ1S19v L9fbV9/WHsAeyVvt8eKMhm/g9F5udO/Vo8EkESp6yPp2z4UVe9v6O9YKdaulutbZE0fnajiJ 2/P4vmNnO7bW3BXhLjYSY2VZw7e39CRk0rWhprmydqi7P8vD7n3B/qqNrb6mldKXb/v8M9uN e+v4+rLHr4wOeiTPnpH9wePm9qHTz375XmOAe7mnduvOga1tJ6f3ZUcD8uY+3Ik4mF64KQaR mVHYPdA3vKurvW2j6Cnrbyov5yqRk/7AyorqskGuyd9R7kFmtgTa1tWLUt1ebheXrdwvitwx l5/6Dq5tF7fbQ0miHDVQkpCPUppWIwvIDuWmr3frFsYMqvCMM+ytx07WIz1aJ9XftWNi5+kH +H0vxJ4e7KiuHWxv2NI5OVTnrV0t1fOrTmSCx42ae5pqqyr7fEfP3d/1HW/vhsZvCU9WhNr7 5TW8Z7NUL02fPRw9ds+9E70ntnUP34jskoeGsiN1Ql3Z2IWx8MHWtpvW3QF543rf0MPKfE1L u6+prNJzb0eLut08Ve9s2K34uc4NosQJV7mLWI4Sd4m7H2D+tPMUnoCopw7LdYVY3PT5Glhy IdkfL4hX5k/juwcL8qHiCOd8P6gYemVq5c5PgS9bOhcW5ueh/KRERTq5MA/tHQAoX78w/4/u 8pPMUuklSG6X8Bk8KD6H+wptCnAOVxS9PFCJ9tagKQnXrsBG13G7C3Z6XHT0LmKLc2d5cS9y ZB5qIK8voM5hVxZhBYy5soSt065MZ4ZduQznXnLlcujEXcWRK4R23IEceQWsEfMYKlFeQ5EL 5diKiLIrc1AhXnRlD1SLl1yZByJedWUBdX7uyiLa+a0rS9j/oSt7QRc/ceUyINK0K5fDaekh V64oOyz90ZVXgLwyj6ES5YmQkcqaWixuk56urs3kmBaOK2aEHAiQg5oejstklj4GbDOt2uqZ QNhIsrPCMTWWTijmsKHbhww9pprbWTe7HVdNSzN00h3oCtF2yzhOHlcnWohmEYWYakyzbNVU I8Q2lYiaVMxZYkSXuXZnkTEjamcUU10OTYmiGbLX0JVEhAxht562sVu1iZIIjMdVkjcRMpKp NNU9ZERUUydRhI1QTJUYaTuh6SqZczBbFIhrkFlrt5ZNPjQ0Pjh85PC4f0axMAbHVtQwiU09 ojc7m1It1bY1PUasLIaadKERZ0YcaYiolhaj7tCeoRt0Ctk2KNPDmj+ixTRbSZCoElYtkolj 3CSDfJE0nW0bBI0zZ1HNtGwyZyTSSZUhp5ajStJIW2TGMGZZH1Xs7u/rYr6mKCuDpk0HCnGN mkbMVDBTemwqQMh+ndkxWWKUmQSdvq2XGZTJ1FI2xhE5Azolu0GGMbMaUjKimklF18lJJRWV ySHFRiQZElJMnC0Xsj2GGYgrCXxHHFZVXVNJUEsklIySUGQSQo0ERhbUYmrCyKADE4GNGAkM F7OvR0gLjU7RDT3LgjawkjVMnksvosQ4S/htIQoNzEphpjUaGU2bomdJUjU1m2g6Y8vNDVJx Iq6hUpJqhOkUjJhqWGoxf2nGZVSx4lg+Dih1TsUSM9PJVFYuVJGd5wmPmnMqmcFgF+Uzqpgk aWCSbTUc17UwDcIIY2JpdSFDOsNpoHeTWHY2oSK8vO18RRZQMbAWQ5ZCP5ZtGrPomYKLIr3o 1YorqUJxGVj9CSWF6B0IlpHCCrARhI2qdBauN8y7YjPSHC+oGlEThPWbThwIk/lkyw7jRYi0 3iJaNKqF0wmb4nIXGlV31x6iwKykVBOTkURDM1lnfdPAVFuJGjEKzHJ3AXdFscylEoqOtNFo cX8K2xoSG1ONpGqbWSSXrT8nzogaRU/MdlB9QEMOw2lzTsUcx207tb2zM5PJBEr2uNv3QghP uyk8K5qgQQziYAPBN0wXfjajdAx7w9ir4HgE2wcggPeD2Kuzfhlbs4XWAM428fys4lOFM6gb ZmfpMH6TqNWH9lT0koYEsziMozrqHmLPGI6ZsL1EuygdZ2MWylSTQDfa7kLs+fEWGHc9j+N3 AtsE+y28K/g1mVfatpkdlcVCsSooqWiDopnFPgOiXyPqxb4IvmHpPBvP2wqz/nVYU9gMiobg qZxGpSArVHvI1dbRh+1qqywvVCOAXuPYJstQhBjXKTbLsUt5jTCZchZ12XZYMZkNg2knGDLa nlvEs1VgZDHCIrZ2pvNVng9hTOMwiPk+gieTcfDj/1YKznTyUIqLyibLTT5GJzYbKzSFssWY sBneGPZb2O9kNbmENbLIR9ythgizQSs9H53uotXxm/dCYBuipbynGE9+nKexCrIZBxSngplU mY0Msx5mPjNufRFEYRXqzGBInRzmI4viHJNhp6wbaDWNMaglnOcxU19JlimLRWXgZ7ZEL2+x m/1v2lUS11ShVgbRl12YsTxfo3g3MEKT+Uq6/E4xTgnsZ9byeMySFaMgnkTB+zboLUEoMwRf VRvjLudFRqfYzNJMht01q7lVMsJmUww6s3ESpRRakJl1hVUIjTvDYlVY7HSGfJu1PeaugTjL bILpHEZtFS1rLLIgPhNsx8owHYXphFwbCTdnQVYhKrYN1HMiMF3GRlh+new6a19nvlsKuVNY /enIRTHThrsna+7KW1y9DpdOPm9fvy3Mbj5jFo44a1or5Cy/2hTmmTCEJmOZMK/F2lq8bpyq OMEqwrGULNgIF7xEXIyODes2dvL7Tz6OKMMbd3efUqZUtjM5u5jJVgp9Z8m32YvsZfVEWKbm mP8ZN7N3Xp9RllXC8uCsZFo7tGJ0Vjn5TBgoOys2v3c5NaSX8Gm4sZvMPkVG2XLYW4p76R65 nKsis1YJZyk3Hou9z+i6U90dXnd3V6d6nVgthjF1m53LcPf+BBuX3TdnkQWLnRPirneHCdu1 mvflvN+c9a4wrXyllcbiWI2wtUJK9M1F+XDYLMZZfNs5+XVYHC/Up4ZaUYaLni7sAl9zS04O jvXF7z2HC2etpJi+szKSLqIZls3i+zufMZW9D6Js58wzZi05Cyx+RxXXXIoxrbvVls+tc34K M+6ciqV7isGQ0LGsW7nF919pPiOs7p2YiriD2HqA7WWOpzQ+59iMAFsbdLdM4bmrEz8Z9gnc 4Rz37+jS3yRg4RP4V5fH/UVhlfM7CLeO/QLyn788pQ0e/usuwUPY7zjsl6CPFxYW8A70DvQn 1f+Fy3PHER7+f33l9U88xHNrZW5kc3RyZWFtCmVuZG9iagozMSAwIG9iaiA8PAovVHlwZSAv Rm9udAovU3VidHlwZSAvVHJ1ZVR5cGUKL0ZpcnN0Q2hhciAxMTAKL0xhc3RDaGFyIDExMAov V2lkdGhzIDMyIDAgUgovQmFzZUZvbnQgL2NtbWk3Ci9Gb250RGVzY3JpcHRvciAyOSAwIFIK Pj4gZW5kb2JqCjI5IDAgb2JqIDw8Ci9Bc2NlbnQgNzA0Ci9DYXBIZWlnaHQgNzAzCi9EZXNj ZW50IC0yMTMKL0ZvbnROYW1lIC9jbW1pNwovSXRhbGljQW5nbGUgNjU1MjIKL1N0ZW1WIDAK L1hIZWlnaHQgNDQwCi9Gb250QkJveCBbMCAtMjUwIDExNjggNzUwXQovRmxhZ3MgNAovRm9u dEZpbGUyIDMwIDAgUgo+PiBlbmRvYmoKMzIgMCBvYmoKWzcwNiBdCmVuZG9iagoyNyAwIG9i aiA8PAovTGVuZ3RoMSA0NTM2Ci9MZW5ndGggMjg3MiAgICAgIAovRmlsdGVyIC9GbGF0ZURl Y29kZQo+PgpzdHJlYW0KeNrNF2tMm9f13Ps9/MAGPrDBQIDPODYB29jY2GBeNjY4jwaaEuoA oQnmYeOMhwOkaR7LY2nSBTVtl65dh6KsU9M0i6rIyaq1irrs0W1JV2Wvqmo3baqqKaoiVVM0 aVUriNn5PrtJ2qnVpO1Hr33vPY97zz33nHPPvR8QAMiFQ8DA8P2bHa7C5pPlADmvIzWyKbSx 7+mf7XwT4Y+wekenosnlwY9ncMJ+AKZ59OF5cSC5Ffl0GYCwsWR8igq/jCK8Gut78ck9sRtD xy4DsGsAFPsnxqNjsXfWLQFodJK8CSRo47wR8Y2Ir56Ymn/E+iD7CuJJxG2TM6NR1XuaMOpz BnHDVPSRJKmSdNGcQlycjk6NH/se8wziqC85nZyZm8d9SBsakvjJ2fHk2pG/IJ+ViGsA93oI AXqeeV4icGcBRqdm651wt6xgaVxZSWewpaUAtoEiva6wQMjPy9VqctQqpYLnWIYSsF1CcSFx YWTBFIxF7LZLtQChvkdSYO0v9g3IXavcpdS1AykSyv8CyW6z4j9Fa+22TENGUmAKXiSMOSiG xYnoWIoxYy+m2JApKo5dDAQW+vqNWC6CcdXApn4JYM1ieCKlMHdFU/BAv7EsxYWaBsqiYwOb +00O06dlDtNNXNv65dKjY186kjMH/Tf/Y7L/U5kkFBT77LZf4LbRriFRUnKhK8WEoonhzuGg 3dYlphhT8BKhXOjhkCm0YEJGijUFDakcaa8ozdCVSAWioqx3qqMMYQlMhctS6wYGUpwpiCKI KZhIUVMwpak12G1i10Sn3XYZBOBWfh5EJS4JHEi9KF6kzELQlCLRYBgNFujrTwWGB6T17DZk hYQQqlsPdlud3fZru82f0tXagQclxn41eKA7HUnPrbyNUT3FPM9NcWfJ62QCI+3P8C78Ea7D OXgeTsMpWITvwJNwBA7DHtgNUzAGw7ANItALPdAJQfBDG/igESW6oR4cYAcryjcqlhUfcTe4 N7gL3CluD7eb28Vt4pTs+yzPsswy8yFTQJfpP+kH9Bwdob3USCtpCfmAnCHHyT6yiwyREvi7 rMvbsjZX4Qq8DpfhJ/AKXIIUvAznUcMzd3R8Dp5GLU/AcXgMNT2Imu6ACYihtkOoaUZHt6yd MVChcnI1WrNGYJ3JXK3qQG5SOc4Pr9nKJZMebSeceI1pPVputQKVEgWHPzxhCqgKlDVQHni2 AVSgoOxhBlmHFQR4qspjrPVgdRoFo1CNleDpMy4HOVgCI3tlKYjo3MoJ3sMD1EEA+lCvHYGo qVLPtKpdxNiwuYKObmVUHBFIfV6riq9XG4F3NtQCXx8eBb7UQJ1bNxE8gnqBbn+wopQ2Oaiq kBy4bzvLVZLNZO7BDoE3kQPmJsKvInMOyOfZCdSpFpWqdRY2NLYTt6u4AhtvoxchPNwKwpuq qi2ehkYZQAqvQMBSbWnUFRUXuV2NXk9DtcVUpeB5khFQJM/LJdlRxRX03lFzkZoZ/8aOzdd2 7PlDp7YoR5PPL91QG/KUBTxz9uBzL560r7b0JNt21kQOCqWenj5/MNgW6faWCowGZ3a0RMWN G03DPv+UPXLQLlhcpX2Bjo6OzQ+47OzIy7t9Qf+m3kSsgM8xaFWfvi8oNLgCl+zuHzl2/zP+ /Z1Nuy5sOdA4FksunppMxBq/uTz70t6mUNiz+/z53U1dwZa9L13qN9aWxidPLU4m4hFMnegD YJ7lIuCEUMDnKClmdKDS1Vg0oOSImkCxQiWglyvVDpK/SlVLeJZu4/JVXBVxFquZ/lVKhqcu tDE63m+1Os1io7fRzRglg7YznoY6NBOvQGNLlkWLeRokE1YQvS6XKhTkHdKrNpQ0NN9Ws8Tb s73f698T2b4o6nKs/kCtrSOWGj9wNdk78zf+o9Oash+kT1lZJRmM7Dz+/e6hxcHd21ify9Xk qx9PxeKv7nzoh/t/xd8AihEOzBvcFJRBBXgC1lUltJjUsLyaJ2x5nkC0Kj3RVJTwTIGWMKSA lJOEsricI6K0C7+1rs7qRBX1gltv0uswBPSSXz3o4jZi9BAvIxD6lq74+nWjeHTh8W8/edRk vP0B3aceSgvcVPFym/8tV47Wz/N+bY7rLT9zXbtsoZ/cVkqWXnlbtvR9aOkWj9vOdDaptJQa iF5RJPDrS2ins0zNKpusFjxYDFuQT4fFdsI5iaeWcpTuKwE9T7szZ8wqhbNLjlBvJiqlmKQY woL3TogiWkc8DXesX50J/XakWaolzyh4K6ny/F7HsHThuxtOrKMsVeSy1n1PdA9tmTi96fHG mUdrYjNt4z8aba2x+lrX6PIqz+5o+YbJXJpbWdrnfG2Q16gMnL+QJSQoVhGiUWrZtkFnWXpm x9rAo6/+9LGatsr4j8edXk99nZePfavV7X3iofjhZnf7M9pi9FQELXIWLWJAmwQDvjri9Kmc dUaMvTXES6Dep9YptdawbA4lsbKqOjLs1xFum9kKvJdsZ0sIz3Rnog9NQqpyib4CHVqUPeTt VNqphcOgQzgTd5ko9LotUmDmEkXmSMsc2WI43kLVPTtbCFGVrprsCDxkKi7i8ss7b79rDHtc 80+t37Ar5LVUN7V6XMf6Ttrd5dR8eHDrcX/Q0xSyD+xoGznr7rAxLWPnRnIKcpXnwuu3eJot gwMb54+4Zl/c+vILXYEhjye+1uZ22xz2vkTvwLuL2yYDXeufHRt/br2zuTwWmVlC28yRfO4N egu0GMdioMTM5gBvLijBNMvqSS5pV4CaI+XZjFuIGzJmvGyssgifw+YMApvOKy3NW6KCoaiQ TQuGDEyvSlTBYEgfwUZARMoHt9JA1ZihFfKqlMGr0swpQElZBUP3cqgV+SzPFwpGvQnPyS1S eu1aGm/U3k/O9Eoy7Gkg4TsyOAq8wgyyDBboXtzBPTLcgsmDF4b96lVSkv4w3asY6v0E5Ni4 ydxinsYbtRPCgTYPcdlUrnq1GpQMMSDSpuPFQtrkNKnyMTrKVQdKioBrIyKZzrMzPBAPTxNB i8AzYXklKTmRe1KQ5HFBJx2fNuLGbCSnc7nF4MAj0ihHjhQjrqJCVztpI1K06HVFpKfJYvEN nOg5GXc0V5tNloa6bkdHP2N8YsvRN0d6tvx2p1jRviW82+RoHgxOB16Z83XuoCmr221dH2vo s0y1pD2OYNBx5Iiz1zkWfmBsevL+x9bYLcxg/ZBxbZ3Ncbs5NB/8q6pr9Xxo7ThevGb8/YNs IS9QG71MP2YMTDNzBQhzkzyFVzJPFslOgHQk07MvQIwWoPlyuLsPWyYfvlDwaXtrhTubxguA R1MfusshmRo9+Ke92/Na/wWM8otzYeUCujqfl0Bp8p15KC8IoPStXFhZVObLku4tOXyWxJ+H OXyVx7D6sn2E98Ec9cEtrHYJl4fm4CtLfr7jKhpcq0j+GnkYaRI3H35zZ43KrOZSyyFGsrMU sDoLMzjemoVZHNOShTnkBD5TDLFQFlZgXYc8wqoQar8DE1DDsSxM8QX5VBZmQMQXWAZmccyV LIwegmtZmEf677KwJP+90Exyz2wiPjEv1vt8HrE3MToRnR0TN9SJ9yWmRyfkD5Te8fiuyehs eGZ6vntmOj4+2yyT5cZZ5wzJHzEhmIEkvktnIQFxfO/Nozb1eA/68DUq4rsvAaNIjSJ/DPEN +PoSMeMmYFqmj6KdZ3G8E0eO4/xdMCmPDaPUaZTVLfdx5M1C8z2j70JOlChp8RkOmW+xryw0 a2ldZiyRkg8P/12h9yIMsFx2ouIrZylB8qUavt6FpaIco3KU38JvUmxBaoFkT8P/q9Av5TDZ U/J5fyj+h7WkNKLKwl93D3zNy78BxgXknGVuZHN0cmVhbQplbmRvYmoKMjggMCBvYmogPDwK L1R5cGUgL0ZvbnQKL1N1YnR5cGUgL1RydWVUeXBlCi9GaXJzdENoYXIgMzcKL0xhc3RDaGFy IDkzCi9XaWR0aHMgMzMgMCBSCi9CYXNlRm9udCAvY21yMTAKL0ZvbnREZXNjcmlwdG9yIDI2 IDAgUgo+PiBlbmRvYmoKMjYgMCBvYmogPDwKL0FzY2VudCA3MDYKL0NhcEhlaWdodCA3MDYK L0Rlc2NlbnQgLTIxNwovRm9udE5hbWUgL2NtcjEwCi9JdGFsaWNBbmdsZSAwCi9TdGVtViAw Ci9YSGVpZ2h0IDQ1MwovRm9udEJCb3ggWy00MCAtMjUwIDEwMDcgNzUwXQovRmxhZ3MgNAov Rm9udEZpbGUyIDI3IDAgUgo+PiBlbmRvYmoKMzMgMCBvYmoKWzgzMyAwIDAgMCAwIDAgMCAw IDAgMCAwIDAgMCA1MDAgMCA1MDAgNTAwIDUwMCAwIDAgMCAwIDAgMCA3NzggMCAwIDAgMCAw IDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDI3OCAw IDI3OCBdCmVuZG9iagoyMSAwIG9iaiA8PAovTGVuZ3RoMSAyOTUyCi9MZW5ndGggMTU0OSAg ICAgIAovRmlsdGVyIC9GbGF0ZURlY29kZQo+PgpzdHJlYW0KeNrtVl1MHFUU/u6dn11+CsxC 2Sm0OMsKtOwuhVKgCwXWnbJp2hqRIoK2kekuZUmApQutRUOoaZpGarUao41p2gcbojw0U2KM McaoTavxwQd9oI0aNf7FxGjbRNMK6XpmdwNaf2qiJj54Zu+93/m555x775w7CwYgBwcgoPeu 7WvXqW899iIgHSdpV7u+rfPp72N7CH9LrT48ZIz8WHN1M2DfBgiN4X1j2r3fd64A+ALpX9s9 0j/ElbcMgN1O7WL/4Phub+XhHWT7EfFmtM+I9L+rHyD/ly1/URLk3L/wNiCvJv726NDY/gI/ P0Y8+UfBYCxs4B3WTvaPEZ89ZOwfYTFWT/pe4rVhY6hv5/vvfUo8+WQzI7HRMVoHkd1r6Ufi fSM1ocR6iv+K5R+0VjIEn0GSpGkgPDQ6XlONRUoQNSQSN1Lc/HyA+kDh8oJ8h5KXm7MsOysz w26TJVHgDN5ZcqdrU7um3MHdXT7vbCWgd+434el2+nuSw8bkYGZW9phMz7tJ5PN66GfySp83 1bFdJtzBs0woC2ohLWpETKGMRs0UdbehRc4GAlOd3S6is3Ct7GnvtoBYpoWipq2szTBxd7er 2JT0DT3FRqRne7d7rft68Vr3NxTb88fejcgfWkplwdZvfjO59XpSpDicfp/3TVo27auuWUlO tZmCbgz0buoN+rxtmim4g7OMS/o+3a1PuUlhiu6gamZZayVvatuAGTC0ZN7mHcWELWiGis3N PT2m5A6SC+YODpjcHTSzK1WfV2uLbvJ5X4UCKfFGkJKYVSRYo6ad5cJU0G0yIxiiDQt0dpuB 3h4rns9LKl3RKd0a+LxVPu95n7fVLKj0QaYnGyuxGvXWiSdm6Z14UJpmHF/jEl7DCziBJ/Eo JjGBvRjFHkSwA93oQic6EEQrmtEEP2qxFj647C/bc2x7bE55Qm6WK+Q86UtpTjondUjtUkj8 TPxEjIlRYUH4UPhAcAoOfpV9yM6xN1mU4l3C23gJJs5ghqKexnGKehiH0nGjyXhL0ShWoMhe LpfkqEJJV/Yy4X6xW+5as1ns6lq9bCeOviJsPLTK4wG3iluih6rChtJAcQWnJYsVxNi4uFMg 1U4bg8wzcgVPDTzVLsWlVFBjEOFaCEqYh0t8fT5IJzyKUeFR4SiWIz+QlSc1Z5Fn5kzOq6qq zl/fwmvXFTpL2PKCHO4uLa8oLx9tjD3V3v5UQ0NT03ASbGjkaD8Wb2ryN1j8cHNzg5+KnWEm cUacFjuwBqsDrhLnKsglpWsgZ5eINtgFsbyIl4I32pUMiXnSmeaX8FrXOipM2e0qreJ1VvxF PofZ0vhG5rNXjrBdilqUe+Pk5GdnwpGZrybYrtwiiz9y5dlMYdLSsbmDcxOqklt03+lLDz/0 8akdRbmKOjF3UFUUNbl2iDHpKApp31eU2Oi+YWK+LTtTZI5Cdk+enUtMpbw8ra2eauYqLa9b X1/rUqijJBTKgiRi8cIjiqoqwmRhfr66wqGo/KQVkI8pqtPx03mHU1XosMro2Y7ncRFMDLFj dHQye45ZNzBSI/8Au7mDxixp6dIS8nAT0bV1OSFNX+ugXGttZw4saViqPRF/vP6B3I0/QLDf PBeJ89c67N/JtQRl8KV55K8FyDiROP/T1/bvkp5+SZKcFonAqNSFGfFz2rnUK/g4UulyqjeZ 9hGsBafI0NLm4cKir9vSGVq9TBxLz7LRvqSwQMXvXQwkU02ksEQaPY2tug6lMZ0WtpGOiRmE WhYxo0ym0pjTF/CZNBZQSvWXwiLZXEhjOgm8n8bWrXExjS3/X+ixkfH4QH90TKvx++u0joFw 1IhHtC1V2taB4XA09ZHp6OvfO2jEQ7HhsTtjw/198caUPNVXV1XrqW+RjhhGMI44BtBPtT8G DTVU937UEeogaZikBukjxG9BFfVbSTqclIcxRLs+TjOqybaPPOzFYNI6RH6HydudybGfdHE0 /sr+l7ia/Fq5LEmQ+rb+KfH0rhekbFkRNfl3rG5Jtw71KxLx3yaRa8ksk2/8ZfqPQT2sHuwv 5M7/kRyEf3eF+J/+Bv0MnnypfGVuZHN0cmVhbQplbmRvYmoKMjIgMCBvYmogPDwKL1R5cGUg L0ZvbnQKL1N1YnR5cGUgL1RydWVUeXBlCi9GaXJzdENoYXIgMTUKL0xhc3RDaGFyIDUwCi9X aWR0aHMgMzQgMCBSCi9CYXNlRm9udCAvY21zeTEwCi9Gb250RGVzY3JpcHRvciAyMCAwIFIK Pj4gZW5kb2JqCjIwIDAgb2JqIDw8Ci9Bc2NlbnQgNzA2Ci9DYXBIZWlnaHQgNzA2Ci9EZXNj ZW50IC05NjkKL0ZvbnROYW1lIC9jbXN5MTAKL0l0YWxpY0FuZ2xlIDAKL1N0ZW1WIDAKL1hI ZWlnaHQgMAovRm9udEJCb3ggWy0yOCAtOTU5IDExMTYgNzc0XQovRmxhZ3MgNAovRm9udEZp bGUyIDIxIDAgUgo+PiBlbmRvYmoKMzQgMCBvYmoKWzUwMCAwIDAgMCAwIDAgMCAwIDAgMCAw IDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDY2NyBd CmVuZG9iagoxOCAwIG9iaiA8PAovTGVuZ3RoMSA4MTk2Ci9MZW5ndGggNDMxMyAgICAgIAov RmlsdGVyIC9GbGF0ZURlY29kZQo+PgpzdHJlYW0KeNrNWAlsHNd5/ufcg/dNkxL1uCuKMrkH D/ESJR67okRRoiRLK4pSJYvD3dkj2t1Zz8ySolXXsh3HtdxUjqM0PpLAbVojSFNhrRaBYRTO 0QaumyYtYBRIL7QFCrQKjNZImwK+yP7vzexyKcmNixaFh9yZ/733v///v+//33uzCxwAVMNV EGDp+KngQMPAc6MA7n/A3oUT4aORG/+W/inK7+BnOJpRcjD+ngKw7U8AhL3RFZN8wbiRAOA/ BOBG47lEhq/7Po5zO/Hzk0R6LR77fY8AIJ4HkLNJVYklhmY/AKgYpfaS2FH1JvwNtnPY3pnM mJd3fEm4ge0btJ3Woor4pvwsxvNn2G7MKJdzXDv8O46/hm2SVTLqP7/b+lls/wSg9sc5zTA3 fo4jsO0bdDynq7m3f+fM4+h/D7Z3AWK9igL/TeFpqiW9AhDNZFL9fVC6NvAa2dhYt1offDCF 96nmpsaG+rramuqqygq3y+mQJVHgOfDdQnNhcm35mjcUX/D7bvUAhCOXC9C72DJ2lj32sUfB 3XO2wIVr7+jy+3rxv8D3+H3WjVsugDf0Kid0hchBklRiBaELn6Qghr0Kib06NXUtstiJ16vQ ue3siUUqiF3kYLLg6JpRCvDAYmd7QQqPnm1XYmdPLXqD3vfag97b6Lv3460rsY/VlLpCk7fv mjz5Huuqq28Z8/u+h7CR1zChQV6bKQhhJbV0YCnk982QguAN3eJ4KbwS9oaveXGgIHpDrYUK ihWttc6kClMKYXEXpttRpmLhYHth9uzZguQNoQnOG0oVeG+oUNnT6veRmeQBv+91qANp47sh DOJWnQT0ScirvHAt5C1wSuggEjYVWSxMLZ2l/vw+HArXhTHcfvD7An7fD/y+yUJjjx9kcGLt d8EgpNbPrRsbb2NVPy08I2WkVzg3/Af8FfwAvg3fgN+C5+E34IvwBXgGnoan4DF4BFYhA0vw IJyHcxCBk3AUZuEATMJ+GIMRGEKbQfBDL3Q6wfGuo13+ofy6/Jr8vHxdel+6Lf259EPpLek1 6ZZ0U3pEykuadEKak9ziP4p/KT4mXhHPiQviDuFD4X3hP4V3hb8X3hReF74tFIRnhCeEuDAp jAjtQgv/r3w3v4PLcPdzToz1LXgTvgdvwB/C6xj1H8AtKMA34RX4bXgZvgZfgRdLKK7D520k T8ITiOZRxHMF1sCAhxBVEmKwDIsMFcUUQkzjiIpi6kdUnVPbHEHpfkdXfYcYvFJdVf18zRMV V6Y18cqVwaoD8PnXhH1Pbu/tpQuNbi4S/uHu4gDPVPsAL4MsDoALHLz4koBDLzk4kHlXjdDb D719nXWddd344UCEzg9DEnwAneIbH4Swwm7ijvIUoqyA1qk6V62AJjgxjimUuCp7dlej7PXs GtozPDjQzD01EfBPTfkDExnrMUFtrBvc19BGO5Cp1jZOrgUXh1ZqHI0C/6KrTZS47czUJEZi GergmhqrOQfaDXLYMTLB7edkrr1/377+0PTuo6O4FzhFUZx/YAf1wT187vCvPzoiijLvltxz ryc7u4CHkxu3hXXccXohDENTgY5tdcJws8vJVXLDQcktOb2cj+vimoPE1YasDHQ01oj1VyZ7 ENgMRoNXX8MwOkZQLdajqdEhyE2NgwPDI8Pj3EATj6i7KfRuDJGjvSMNKHs9sgNjr0YILc3N sYGnJ184MrHcJjsC7a2d/PXq/aOHvEKdu6nnd+fi7id/de6yt5erH9ID27bpcqUvHlzq9Q12 NNRXVO7n3vjOE5Nzh/dt3+bd3lYpf/Td/d3nH3+uun79+2fCL85PH0/NrH/9wJ6OHu5Hu7Yb E2P7H8+dinbf3z24Hzkf27gtXkf0czA+NRga9RJZkIfaXC6Ef4gbCla7q537uLZgt4tH8DOh ydG+ntb6Stcxr8xJ3FErs8iA3MthRiwWML8tlATZUcZCc1M1z7KP6IeH9nQHOIpfqOa8bN5Q A3Zij5xxVrmdVc3quRNJ6Vf+6Pqot+7++fbgbL5DqK10ut3H8pNzS7WRzqYK94An9/XJ63Kg 7T4Px71UNdE13MvP1lU5v/qyMn/z/J4jY/te/uNDw9MvXKl2O0/fOL00u7vnr5+8ENi9c4fn 4asfvbXdS1pFh4NP7Ok7v9ewmBDWkIkhCEztpPgHuM5gDeLnxBZht6vP37tzx7a2Ruf5egp9 xKrEQOAO8BPcCC1KhwWMVuTIrjKke0buAnok7h557OTsg6J7cifxSa6xnR2jC0+Ofgnr4D7C v1BTBmwhPH+E42fHXW3dwV2koqrbz7dr2Y++1eEhrW4nn+gfpFAQyzhmdQixKHB06sADx+cO jQ57SXurLFR2BV1uroZ7kOsK1iO2Se4UF+KCwXFW3GcfOD5/aHLU3+slHa31NXLtA80ziDV6 Z5p5itTCyt8Lq0xLenAA22wB0PHuXd2batYikR3NtDoaSuUgf0uoqXJVVKrPHh5XO5rcOx+a GDu9R3Tt39nZy7uGPF2H9n+m+Xyju2123Ig0CTWVDbVvPvRIuNHVaUye0sXKfV2eHr7lqG96 4jPNFxrdlcOeveYCo27hudMDgXbPzBmOJ/vub67oDnZ11jq7fUK7fmb92aMju1sPHblyrLr+ 0e8o2zrDv3R4v7u1x9dN6pxH3W3505y22NvZ4dl15Crldmz9nHAbuV2AI1Nhz31NNbhiIi1D 3MwE79rORYJuN+cMc31cS3DG1egYH+UngrsdtQ5BPDkXHh8a7MUpDTVVVYuEltGiRW3PFmqL a4dyONJQ2lOqBW+AK+MP+UTaOjhK9CRX3GBGJvgRq8SsCqs8Zk7OX6hbqL+vpi3yueWvWmX1 p01LTzx/4sIvy0gakoulJzml2sjI3DOp9trF/Fxuh7fB2da7HZeewRg8e/30hSPe/n86fubQ YNf8b/JXOjzb2yvlD98+/eX8NB871oKE7dpR597Vx3HdQfX3UqELg3vGIqEZz8nLR4/3TT7P Xti68O8QPM5x3Oe4d/gXhRHgROCexSNI5l7kHgJYP2c9xS9DnK/HI6pC2nz5E2rhjgtf/97d kF5ZP4fvr7iVX90c4ayP/8cfXbtYs+/nIDjvnAsbb6yDc0mmIp1cmof2DgK+emy88f4zziVm qfxyyXYXPwY36Ucch5Pij2BM+AqMiV+EceEdPI2pigtW8GRlqlCJPppR+hcw8fCko23cRMn2 gB0xvcvYsh3gqTxqywK+VE3asog6J2xZwhN3wZbx8MYVb8kOfMtJ2LITavENw5Jd+IZw05bd Yjf8nS1XQLMk2nIlym00ctGFrZjUZ8scVEo3bJmHaullWxbAI920ZRF1/sKWJbTzt7YsY//P bNkBWWndlp1A5LQtu+BB+dds2e08Jv/UlivAVzNqy5Uox8Jabk1PJZImGejrGyQnU9GkosfI XIAcSWWjSR+5RB9Tpp5XTfVyIKplrO8RJ9VEPq3oB7WsOa9lE6q+1+q37guqbqS0LOkP9IVZ hyeCBiLqooekDKIQXU2kDFPV1RgxdSWmZhT9EtHid7m3Z5FTWtxcVXT17vCUOJohB7Ssko6R GezO5k3sVk2ipAORpEqKJsJaJpenuvNaTNWzJI6RYyi6SrS8mU5lVbJiBW3QQGyDzFq3cdfk +ZnI9MHjxyL+ZcVADJatuKYTk3pEb+ZaTjVU00xlE8RYQ6gZOzRizUgiDTHVSCWoO7SnZTU6 hYxO++iXOX8slUiZSprElahqkNUk4iaryBfJ09mmRtA4cxZP6YZJVrR0PqOyyKnluJLR8gZZ 1rRLrI8q9o+N9DFfFykr07pJB0q4TuhaQlcwVdnExQAhh7PMjs4Soyyn6fTRYWbQRy7eyUYE I2eBXvTZIKOY2RRSMqvqGSWbJeeUXNxH5hUTI1klYUXH2b5Stk9hBpJKOu0jx1Q1m1JJKJVO K6tKWvGRMGqkEVkolVDT2io60DGwWS2NcDH72RjxUHRKVsuuMdAaVnMKk2fTi1EizjJ+PUSh wIwcZjpFkdG0Kdk1klH1lElSWcaWnRuk4kwyhUoZqhGlUxAx1TDUzfzlGZdxxUhi+VhBqSsq lpiez+TWfKUqMos84VfRFZUsI9gt+YwrOslomGRTjSazqSgFoUUxsbS6kKEsi1ND7zoxzLW0 iuEVbRcrshQVC9ZgkeXQj2Hq2iX0TIOLI73o1UgquVJxaVj9aSWH0VshGFoOK8DEIExUpbNw vWHeFZORZnlB1ZiaJqxft3BgmMwnW3aIF0Ok9RZLxeOpaD5t0rjshUbV7bWHUWBWcqqOycig oeU1a31TYKqpxLUEDcywdwF7RbHM5dJKFmmjaHGPipopJDahahnV1NeQXLb+LJwxNY6emO2Q +nAKOYzm9RUVc5w0zdzeYHB1dTVQts/duxe/2WiQw2+QOqTwVEjiGUTwlOnDv0GUTmJvFHsV HI9hew4CeD+CvVnW78PWpVJrCmfrkAcVnypcRt0oWs/gPYN/KfwO2ocWVfSThzSzeRDHs6g9 z54JHNNh7xb9cnmBjRvYotoE+wLYGy7T8OB3XyuCCH4WsU1wxMC7gh+d+aZtk1lSGSYas4KS ilZoTJewT4P4J0C/1ReBU2yeCatshvqJ2FPYDBoNwW/sFJeC3FDtGVs7iz5MW1tl+aEaAfSa xDa5K4ow4zzHZll2KbsxJlPW4jbnFis6s6Ex7TSLjLZXtjBtlBjZGuFmbN1M5xd5nkdMEZjG rB+HYyj5YRltGXYeyuOiss5yU8RoYTOxUnMoG4wJk8WbwH6D/QZishxuZY1s8ZG0qyHGbNCK L6LL2tFm8VP0QvAda5rxnmM8+XFeilWQyTigcSqYSZXZWGXWo8znql1fBKMwSnWmsUitHBaR xXGOzmKnrGtoNY8Y1DLOizFTXxmWKYOh0vDvUple0WI/+x2rrwzXxVKtTKMvszTj7nydwLuG CHXmK2Pze5FxSuAws1aMRy9bMQrGky55H4Xhsgh9LIJfVBsRm/NNRi+ymeWZjNprNmVXySyb TWPIMhvnUMqhBR+zrrAKobhXGVaFYaczfPdY26fsNZBkmU0znWOoraLlFEMWwmea7VurTEdh OmHbRtrOWYhViIptDfUsBLrN2CzLr5Vda+1nmW9PKXcKq78scrGZac3em1P2yttavRaXVj7v Xb8eZreYMQNHrDWdKuWsuNoU5pmwCHXGMmFeN2tr67qxquIMqwjLUqZkI1ryErNjtGwY97BT 3H+KOOIs3qS9+5QzpbKdydrFdLZS6Nnlu8deZN5VT4RlaoX5X7Yz+/HrM86ySlgerJVMa4dW TJZVTjETGsrWii3uXVYNZcv41GzsOrNPI6NsWezdGfede+TdXG0ya5RxlrPxGOw8o+tOtXf4 rL27WtVrYTVYjLl77Fyavfen2bjPPjk3WTDY+0LS9m4xYdpWi76s881a7wrTKlZaORbLaoyt FVKmr2/Jh8XmJs7N087Kr8VipFSfKdSKs7joO4ZZ4mvljncHy/rWc8/iwlorOaZvrYyMHdEy y+bm+V3MmMrOgzjbOYuMGXe8C2w9ozbXXI4xnbWrrZhb6z0qyrizKpbuKRqLhI6t2ZW7ef6V 5zPG6t7CtBl3CFsPs73M8pTH5wqbEWBrg+6WOXz7CuLfKvsLfMz73P9El/4+ARs/g//u4u1f Fxqt30S4NvzI8L+4eOETqYnlDYn9TMJ+IHHCp+MSecKiZL++vLuxsYF3oHfgtsb+/3jxIPxf IvzYkeKPe+V14IBPS2Y+Jdd/Aa4lMuZlbmRzdHJlYW0KZW5kb2JqCjE5IDAgb2JqIDw8Ci9U eXBlIC9Gb250Ci9TdWJ0eXBlIC9UcnVlVHlwZQovRmlyc3RDaGFyIDU4Ci9MYXN0Q2hhciAx MjEKL1dpZHRocyAzNSAwIFIKL0Jhc2VGb250IC9jbW1pMTAKL0ZvbnREZXNjcmlwdG9yIDE3 IDAgUgo+PiBlbmRvYmoKMTcgMCBvYmogPDwKL0FzY2VudCA3MDYKL0NhcEhlaWdodCA3MDYK L0Rlc2NlbnQgLTIxNQovRm9udE5hbWUgL2NtbWkxMAovSXRhbGljQW5nbGUgNjU1MjIKL1N0 ZW1WIDAKL1hIZWlnaHQgNDQxCi9Gb250QkJveCBbLTMyIC0yNTAgMTA0NiA3NTBdCi9GbGFn cyA0Ci9Gb250RmlsZTIgMTggMCBSCj4+IGVuZG9iagozNSAwIG9iagpbMjc4IDI3OCAwIDAg MCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAg MCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgNjAwIDAgMCAwIDQ1 MSAwIDAgMCAwIDAgNTcyIDQ5MCBdCmVuZG9iagoxMSAwIG9iaiA8PAovTGVuZ3RoMSAxMDI3 MgovTGVuZ3RoIDY2NTcgICAgICAKL0ZpbHRlciAvRmxhdGVEZWNvZGUKPj4Kc3RyZWFtCnja vXp7UFzXmec55777/W7opqEfdEO/oB90N9DQ3TQN4ikQQgiQhEBCEqCnDZKtyFj2KLKMZEuR H4qieDK25VEROaVgVyZxHK/txElmM96MK+PazHhciSeVpLxKbc26XJutZCw1+93bDUKW7PH+ sblN33vuObfPPd/v+32ve0EYIaRCDyEKjfdtrI3oI0+aEXL8GHqH+lt7Bi+UzP4B2v8TvvGd +ycODVw79ncIVf4OIapx55E5+z/IfzqOELmBED6y+9Ce/UT7owloV8L3X/bsO7o7c36HHiH6 WYTK/m1q18Tknp91wZizTJxvCjpUW+h/gPNBOK+c2j93//F9ZjWc3w/nD+07uHPinokZL6zn ITjP7p+4/xDuxCEY/xOc2w9M7N9V970XTyHkKoH7XT10cHZu+Y9InEpcv/3QvbsOfd/0x2m4 /xyc+xDIKk5EfkUtiFcxVxCa3Dk3HQ6h1W0ZtsTycr5w9sknGdhnTEaDXqfVqFVKhVwm8BzL 0BTBKPAyTNdqP73jtCu7eygYeNmHUOvg/UvIP2xuGJEOTdJhSeYbWcKtmk91BQN++FsivmCg sMM7lpAr+xKm3Fl7u31qYnKJcsPRvkS3uibsky9lMqcHhx2wvYQcZSP9w2KDdtvbp5Y4d9vE Etow7LAuMa31I9aJyZGNw65a15+tta7rcG//Z88+MfmZVzLubPr6HT9O/1nq0urMDcHAD0Fs wLXVLi7ydNsS1ToxPZ4bzwYDbfYlypV9GROm9Uirq/W0CwaWaFe2ZEkuygqzlbRNL2Um7NK6 l1qs0BabS+3WpY6RkSXGlYUpsCs7vURc2SWFryQYsLdN5YKBV5EWMctvZmERL2sZJB7t9pcI dTrrWsIT2XYALDM4vJQZHxHvFwzAUKu2FZYbRsFATTDwk2AgvWTwBRGLeOC+B9WhU/nR/Ozy u8DqBeoMs5+5gl/DL+Ap3IXeQ/+Mfo5+iL6DXkDPoHPocXQCzaNZtB9tRaNoGA2hQTSA+tF6 1IO6UA6lUTNqQAkUQ1EURrUoiBzcMW6IG+Bs7Dg7wObYMKtjbjDXmHlmlhljRplhZojpZ9qZ HKNheIYwiM7T1+nf0D+hr9DH6WN0gq6geRpRH1LvU+9R71LvUD+jXqVeoa5RV6mL1AXqcVjz MWqWuoeaotqpNBWjwpSV3CD/QT4g75P3yC/ID8lrZJFcIrNkjPSTNEmSKmIlMsITmhB8A3+A l/AiPoVP4KN4EjuwDf0JXUcfot+j36IPQP6foFfRK+i7gMESuoauAg6X0EV0AT2JzkuInEEL 6CTg8jA6jo6ho+gw4HMQEJpBU2g3oCRik5WQCSMH2JOdjbB+tYMxABKRyyol/z3Fi/LL1KVd F5jLl+uUOfT4K1TTSZvfLxqr6KAY+ICH4pAzY40SFrF0FAmII/SrFAy9ymHEEkFN+cPIH3Jo Hdoq+GJEI8eNLIM+QQ769U+ywNIn87P470GjZciVsSoQ4XVEwBymtZwBw+lXZAzD4HJpnrQ/ 5GZdTk+sLh6NGA2cCnNwWovrEimSiMfxk+6Ez9fXPKy0nt5JK3ia4hp25vSEuFxBV2Tz4Wjw nn9cT9O0ias58vA6XgnuCbBE+C24uxqVZLRyHQUCYPoJFYsYrC2u3a3C0j0TKRyNgNfBr8v6 g64wregLumv3B1pqHZwpmA665IALug9wOcW8Bq1GFMsE/MiAWEyHnTCrvtIuUIxgrLTbzOC6 NHKBARxVHY0ezOCmws18/hCmMGZdHpeTc8YSIGciHqMcFGs0mE0gNHf7UALHkzhi1pukQboi mD9BqTjTzP4HH9TRnGnUbSnJB59XKUpbtAxn3Dm+abMGujvjjZXkR2qT3kQPUybDtIYYyLM3 x/OUmWXNmhKj0o6H8r8tMxq07hJNoc/haaV/ZTL9x/V6ubIEYRxefpedB1+dQtGM31FBEM2b hXpch3lfRKHgzX63RIVGRaODJrTuotYHOGQ0lN+f9oOMzlg8VlcDuLKcinCicOUkGhG1COIY VZgGxKtEzOO6RLwyGjEXBihQBfMTtbz15Pbpfx5KjbcavX7blqNtm795ftpEmUwq92v5Pz/6 5V9mh6ye54/NYtWjb4+f5+SCaoFoOGrU2F2799s7p0da/ubloedm79+WHA71PHJtzGA4/3+e mttz5ntjB57CioO7jixdkRlv/hoIgsGrIPYd5j40jUYynT7bgKjN8bYdY+uoVsEYVPgVXoVO oVGoFLt2jG3dsnnTwIb+rs51rS2ZpsZY0FvlcdptgKNOJecY4mqoncQs3quRuCxt/poaIJio xlhdFShWVCtoMiEp2smxHOXQ3zkkEhLG8GeRQn83vojsxYkSObbSspIHRBpc2TS0qGdp85ec JsU8ic6zMutRDU2XLg5uWtTSRCfc7zQq87+nZeZ5sfvYw/MPa2jKMl9WIj9z8+dn2Nv6sVY+ 3yOP0HsUBhOjMmvMNA07lRDJ8/hPRqWOW+mTKVV89JOvqYxGRrnCLqU8Qmpv/qJEqeVW+mQq QP/i8ruMCli2AfVkst3NSadBx5ZmBQb8Q2lADjwL4SDO+usEHXAtLigVcazrdpKOCsFkUFIv eDo0bLqZ/G1FP82SjSL5/GmJfnEJjVuk8hcI6ZFgBehMFbgcg4NhJag9VRwrXqFPYfF6cVyC WfxFLa4hsTr8gu+RzNNdb3bMcApeNUfUvFmjmD+5c9tWmyyX/v3YLq13IXuhe/+zNTMKgVXN YZXMpNYvXhwf326XZ7O/zQwFaNvlexKpx0/IjPkgr9B0t5x5pr9/ttrZPqXLWr6+tz79rflk pqPthEyH/7ugMHakFy4NjTzgD3Xv12V1/QfbIF3CeDew9SCzHyLeukyqEtCiZXHBj21Y5rMq KB7hNG7GcV9SkANeaZwUjNmWdLKhPhqqDXoqOZqoO20GoGi7RFGAK+QuOJmYBwCQ/K0IhGin Ucda5KqkcaCkwWQux7dTMo6/8R1/pYDNPF9/4CsDnpyfM2m0ak/Ywxvo9focUct0uhJd7RMH EkqFMv8HlrdMD2mBWJe/+cJVkVhptTxEj7ptfIvjaw+YTPWzj+43G/PjMpXhqXVtJS28Iv7J lXILJWNuUQf87/vL1+kOyClNEOXimdqWWMhMc0ahFkcxFw7IGL4cN2I7NoatggXQqI+ZKVp3 WVMJFp6TqCIZpt4j2p7IA/gT3ZRoRSoikgIMtKoQFgqRwSwdjKLpRSMiw4As5IK1/ADPV40G Bj3+UJlBz8kS/q1Ea/IPZttGhWPHz7a4jcZtmabz65KDZsKFrOFz+qaalANreJP28Yow/n68 9HBzPPXl/f0ztZX2aAvWjDw8OZB54vT68baB4W9lU73ZmNLuaLn5+kD50MIpTopC15evU1eZ x5ELJVE6E6NxY41gwI1hk4zma8IOyTVrFVqBxL2CHr9gjdOMgSMalZKmv+UlWpY0QziSAi6A IMq41gm5Qca4KK9JshAxIOtZNWa5Ajqi5Ini5VjmMP5MxpumtTTcn9VeOnBP516byxA6cfJc bvfbZkfCQWTKxoBv1BmIllSxlo2jh3tGM9sot0etYzTgMUymm1+/rzfVNbm+67WF073k22XR X381WxWsy2QjyfMTz3ZVOmbFOP4BSPw6aLsB1WSqYjarpYRiWb3AgqYRrw/bJZHrQ1bQ8XMq N+g4WdQxhtUWAxEuBCJJBhCwAhuK5i8FIkm7klco5gOUQVbvUG+M1+3zRxtK7SptT93eX7Qq 5CytIt3/Pt/3kDusDT3RdqUvOZ5gZY3OMg9Z8vhIdailLvX4jmN9cX+sZW5CQWtV/P77utMD B9Ltbyyku3y9UaOlKuiw8issPgqaNEG2lsjUdq4jjB4smglXgyotOIPLsD5slhjcoOhMJSGo srrLKjtIuH4l2rrvRuJE3Hw3Dq9VdBqcHV7LZgjMLHnnTj5vt755J5lnZTL9jJ6jTW+s0+ff u0VqoP6v7mD1kYq3PsVp/LZHreFECmhY0ye7i+TWAfkBlQ9B26+BtpWoHqUyMbfLaaf1VkEe joK2I0SPWGu4GlRO0XrMGwUXdtrJ3yV4IDZ+3hjQsKRRI/Fb1D/LRQqEFSFiVJRrJRsBEiRW SfApDhBzZ2D2+8e1bGdT3Qye1h64eHnYpZB76iuBAmq+SIHLfX5z4Gz2cl9ipI787/5vWLrb jKaG9sRs55Zq+WMbokFPs3NuQkYKFLi0ZaZz+4uH61t8PREo9/Hy2/lR6hJEv17UmKltwBGs 8clkMh7TndUOSt+iKOUYRU97YywS9rjKtUoZhVRbcmXgvvtWMgxfgd9SXKM4yS37MQUZ19qU FrTKQUATKUAl9C4pVUjiVUrcIjz5mAuXO5s9Dm3M8iNIIIxv7LB5kk7XVDI3RsuSjowQLXPW GyvLBJ2l0qaYfF5DEy3/N6X2aq9ra23bdrjGWeZmkNPLVGWs+X87A8rNs0+99F+aW7d1JOUl 1YFBR5B2VyQc6UNXnl5HiFkjqE4+/VjYG1vXWKrw1Nitgmjrl0TPDjaRRU2ZaDboZ+iMWzDi TNgMNhHC7jAlELAHnSKbjLpskH1p1SoZr9weZHDBtxcdmyhWDN/p2crximsTua9foX4BsbhY a9RViS6DwjautiyMg6sejtY/+6Xn3OvqNRZNdW9Z7bE2ifGOhZ1tA5ohG8ULhBdqKl497rc7 s/TLa3zcw+f1PluwI5n92tt9dWmsAaY/9PqWNm/wD3M6f7LcY7V78h+L/mBx+Tp5gckiH9TH 4A885WVanmWRSTBHa6B0DivAAEySh6fokN/rqTRo1fyJch4kjxeDOmTf7EocL+iXupugoifQ SxBJmeOV3NYKXmdWGdtn2tr7tc9V1D6QIyrBrKp4ZGfbRs3zFW/xTa5mrtmZokafeMBo3HBu 20gu5M23bwyBPIIgytNaHcLf3xhpxhcc4QGnmLGg10Gee0CXlciTKbeFC/WXy6QoQUQHqSrm 8VUtW8pij+TMIFnGK+u7pThYn/uWpxJzOpJ0bNrWMyZgVFANJKegm/x72iZ/yi7WCbjq3Ld3 99odNCoqQaY0gZfJ1gw99lcGcV2AM/sqeJgxlMs0jQ2vT0YBWp72CzYjsejJUI7QYUg0+K09 xB+OS3wbGx5YD8mU1wnjiqaoAiAfX4UcFuu6DXZzIYSaC27XVKxm70o4UVCJb2zBcMVENCHJ LscFDn9D01CTdojqMfs3ZbpHmQe+ndpeWkKlvrajeazUptOlXBW1D0psVFU8KrKx32ESqfjG A6U9LpHEJ+UywhtbXn9wgauxhsmFTGjw5ILRuHV+cnvnYqDGmxiZCNSE1oW7el4ZH/bfYqgn 9N7JgUy1Dfj50b35fz0FxCa6poTLY23Pf3TzNxWubBFNegnQzIEni+RSvmqFjKWdYLN63IQL QGaxMxyUcMyl6kN2wJDf5FOsyUU/G8NbthpfqR9XvFvi85Hq2sLMX2vaXmrVuvoqanMTZZy2 VOVY2NG2UVvZYCxA8oPjd0CyrfNKoCbYkUovvA/svnREJ0Hh92+5tqko+sc3PxBFJ2JeQmtA 8lIUQv6MqxJTGoEK28BKNVKEInQlft5Zgpnn5EHEkoh6Rdi75CSFMOW+xZhC8lUITdSstndt KpLY56trNDvyYUhA+o67w5roE199tpCTkN/XtazJQVpiqbPb5/vjbemu7tSmQ8mON0+eLqYj osex5mdpA0jQhAZEf8uzBGUFLVbjPiy6GxVfgbPhMkEDogRrsZ8zppLhaicPWYipCzF4UMwk /TV3uB3zp1LIqs98xCE6IjMlPeFYzS1rpAoYH16jy9wWYX6+Y29ZpSH05UfONWywjLK8aa8W 6tzde4a3iuGq16CW/w/N7EFecI35h52BsKWKlccDozvs1GhRsxuyT65v7tvesvkHp0+n4/ln SlUqTgU5C8vCTiGro54BFvw4vy5uOQL556MzFwcqHbHUYW+B49SHgJMf+TIOLS5Vk4oiQJhm 3YKvymGzMjMG0RMHCzGo5q6YFONtLC6Fbapg8CDzbcKuG+H3PbbhPinqqpOevnP7nhBJekmz KsnG7NPplMxQ5R9I6io+vvkW0JHshcUX4ic1BCtdh5KZaGs2GY+Gy7UKJDM0CIYwBdRsCHsl ara3ZtMw6iq3aFVIeZgLwNo7V83xtnw4VkyjC5ZYjqGAJqu6XZNoFFLPldIZg1hZ29UvJScS LOQWMS5gkQ829UWiWrrEa90V21DDe3ZMPjlzkeWbHRlZo0u3t2u6JVtCtzRO5gbIuftOSNmy vbrD5qGCiSpPh9JkqbWzJkXzcNxR8Wh319RIiRUw8ARIXSoc3FqijvYozOrJNhfg8ArEnj8x BHnFSFrNI0pwCWYolgWfGtIslxSMCG2wmBClMwtEAcHoRQdrZLFfDEYSpVeDpZRAJLG+UAKu ZFoFfy5FpPdq/nZq/Vb+6f+qpWX6vx8UK73BH4vsfOtpXfLsbkpfCkHpr17b3VtpI0OQFOB+ KOryS9C6ebXXNYR1mqIfPQaa6wRLrIuEKm1qjuGQV3AA02QyzDfgEogdXghIYkHQkWmoqrTp 1UphNMKB6rpvJQBUgXTF/I5ddTKifekL3dTtmf9KyVtwN+ITANDd4kMtIiNdj+yElGCT1qqU xyof3EQofIELWczWC5okZP1SliAWvOlNqh0Pt05ZbHre2qCooEbzHxuNJ14V04T/lTtQ76/y +P9apZWRwQqX3XnzW9nA8MIpSBwg5c81zEcjDV3xEdfGZ52QQBRjCvXvgEVSfAIJOFCoRKjE HgkJgS+BYCI9gIx4HCIE/HS1CEHzfw6BrlC8Ss+YC/6JfL7UZrUq45+9UCdr0efc5WpHqO6N p3KzdtfdpNycCwfcAax41x09kNKVutM6Z9/Yy7/M/ywaaeqNDq2IJ0nHnAfpNqNMpt5ZZgEB WeQX6rG9KGAHqLoL+8OtkqqH1nfEwh64DkRV9DaJoo58AW3XJYqEXS0LV2r3VfHFB7SS/Ldq xbujoFd3Bsc6dLNcpDR8WFCqNGlf/95UQJbWt3jKtM5ybVX/2em2OTuj6Jt9Plt1N3R2hOvC Hv/S5iz+BbirfG31UEXE6/E/9J0X3dH9aV2pJ6VzDdQ93frV9/46GuEsj+zMbAgPrMEsP0pb AbNe1J5JOUTIGFQSj5J0AxHKixbSDE6nJJwWDFxD2MOJWXJPrhkcoFdETr7bwa2WT37f5yG3 YiQSOLfeDYi+UArWafxFbOXYwPOSofy87NQPd/XPsLLGSpuLljU5FG2Rfd90qzbu+XxzeZKc Em3lxjuH3jwy3mWxer22Mk7nCQQi7z/T0H+7yUj1JP0W1JNTaFOmp7ctS/VF9/gFRAcmiNAX blIgfhQPYTneigO+qMzKW/FEmBZowMip5zZlagN2mwnxuq29HTJ6Oxbwi00cZOYzxcwc4FoB Q+/4ojUmueOpSqJYr6ogw3BECg9o1+R29JLoSXrG+Av5qS9WfWrKmkItF8e75zdy0kMYRqgt j3/9qU6732LBVzHBW1XP6pK1KYeY++Eq8DYFH/yfV6aDU4eqE25rNmYwVAfsVmJ2VjNN7WVP nDpDjDePg9NeOKWDDIqEqSr6A6hxylAC1WdqHLgK+ysoSsfI5SymQwqTwlpKvLwQhW5yttKr ZSl8lheftTQUHlOk01Ko4QrpwUpEJSsPqViGvTVUQHBlhCUzhBBP0u/fkAwHFYbemRwrkxHC /hS30rQ3EfD1NY0pKs7sKvSOH6duKDSCV3xdlkvYG7u/Mh+TXph94x/lFm+gMrL5mL1x7r/1 Sn0930ME76a8tAFkK0VxUTYnLsMWBaKInlfIIfEJaMK1NVSFTLDgc5UVmNGLr/TOyhiGJfWF R4ziCxC89qUeu/pOD8S4/XUfWfu6jxz01/t8/U1blfFf8jwlo4Wx4xYW3+dt8IFQWxW6roMt PM/whE3vaNPR9AfegDM8fDTk+903YflGvueVKYMW+iKbj9TUdp09EYHeUjoyeyKnRAi54TOJ HsMIz5EoeZvaQL1Dd9DfYdLMy2yKfY2r5A7xZv4p/o/CgvBr2Q7ZD+Qt8iVFl+JflRdUtOoQ zICp7+KvIAax+BK+B6H8aOFIfxXtJiIt5Mytf3CgNOhTWwahj5aZK/lRhDgvkf5Dorjhwnfb 4kf/tF3d9EdE8Z/+LVr+bh7xNzgvNFnI4Fd/B/O1I8RfWP7uzZ/zN6SZ1m52tthFDqMn4ftb 5mN0H/MKDsNxnj6OLkJ7N21G71PX0HX4flBsf0gOL78N7Uvwm0UiQ6+zZ9AiXL8IfeI1Vmoc LVLtML6IXqF7oH0NLTL9cPyNeB381kbCtA3vlm5vR4uILiwDKWD9Jmjl0XHoE0c16Ker664o oiHuWTgrLh5xoLtCm0JaFCi2abimudhmYKS12Gbh015sc/DtgTFMC9BKrbYxrOR0sU2QCl0o tinkRC8U2zRc89NiG7SO/qnYZqH/X4ptcf7ftR48dPTe6T1Tc/ZwQ0PMPjC9c2ri3kl7V429 e/rAzqnCP78M7NpzeN/Eve0HD8z1HjywZ9e9jYX+wj5UE8oV/kemFR1Eh9BRdC+aRnvAs88B fmHUAJ8YtAagdyf0TsD4JJx3oRrYd0PvAal/EvZzcBaG6nQA7YIZDqN90tXtMO8BGOuVjntg 7F7UeNv1a9shmDcE1f2tHhF7tPwx+ryNFFE3iNeCEi3wZdH/00YKP719oxFTUOzqxt1mI4JM rlCqkBqMTqvTG4zIZC4ptVhR2ZprbLfNWF6B/oIbTeySFBLjP1peXoY9EvcIFy3j/89GClr7 IiuUbGhlu7vOOMR/5u8FJENysAsl2BJCalRwf1qkQ3pggxHaJmRGJRBVLMgKZ2WfMY/tc9ZY jv6iSvtLbv8XWR4LpGVuZHN0cmVhbQplbmRvYmoKMTIgMCBvYmogPDwKL1R5cGUgL0ZvbnQK L1N1YnR5cGUgL1RydWVUeXBlCi9GaXJzdENoYXIgNDQKL0xhc3RDaGFyIDE0OAovV2lkdGhz IDM2IDAgUgovQmFzZUZvbnQgL1RUZGN0aTEwCi9Gb250RGVzY3JpcHRvciAxMCAwIFIKPj4g ZW5kb2JqCjEwIDAgb2JqIDw8Ci9Bc2NlbnQgNzA2Ci9DYXBIZWlnaHQgNzA2Ci9EZXNjZW50 IC0yMTUKL0ZvbnROYW1lIC9UVGRjdGkxMAovSXRhbGljQW5nbGUgNjU1MjIKL1N0ZW1WIDAK L1hIZWlnaHQgNDQxCi9Gb250QkJveCBbLTM2IC0yNTAgMTEyMyA4OTVdCi9GbGFncyA0Ci9G b250RmlsZTIgMTEgMCBSCj4+IGVuZG9iagozNiAwIG9iagpbMzA3IDAgMzA3IDAgMCAwIDAg MCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDc0MyAwIDcxNSAwIDAgMCAwIDc0MyAwIDAg MCAwIDAgMCAwIDAgMCAwIDU2MiA3MTUgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgNTExIDQ2 MCA0NjAgNTExIDQ2MCAzMDcgMCA1MTEgMzA3IDAgMCAyNTUgODE4IDU2MiA1MTEgNTExIDAg NDIyIDQwOSAzMzIgNTM3IDQ2MCA2NjQgMCA0ODUgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDY3 OCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDU2MiA3MTUgXQplbmRvYmoKOCAwIG9iaiA8 PAovTGVuZ3RoMSA3MDU2Ci9MZW5ndGggNDg1NSAgICAgIAovRmlsdGVyIC9GbGF0ZURlY29k ZQo+PgpzdHJlYW0KeNrNWWtUW9eV3ufch3QlAbp68kZCCCH0RELC2AiEBNiIh7FNMI6xLWRe JmCIHxjisbFDqFMnxHaduq6Tpq9MOkmdGZxO0jZt0qYZJ5Oka3VNO2067WQ6rdtMvSbtZGVl ZrW2EbPvley4STuz5ses1SPOved1z9377O98e58LEADIhWPAQHLjFl9A//TZGwDGfmzt64l3 9p6LpExYfgdzePfk4HTFb8P4RH4vALN298wByyy7rweA4jPEPTI9OknF7wxiuQLzj0cn5kbu vZbjAWBxvPYfx4YHh0bfbf8Wzr8szTeGDbkzvBbrv8Z6xdjkgdk9FmUCwKTB+sDE1O5Bz4UA Pm+swrp/cnB2mkThGvaPYd2yd3ByePPRv1rC+v34vqemp/YfQD0w5ddK/dP7hqdzz8OX8P3S 825AyY9hgdEyT0ijuPMAQ7tTszV+uJVWMdWtrqYztevXo3iNmowGvU7U5uXmaNQqQangOZah BNzP4HRxy8nUSVtspM/jfqYaIN47uwyufnP9NvnWIN+WVdXblklc+6Emj9uFf8u02uPOXEhq GWyxS4SxxyxtlrHBoWXGjnfLMhu3DVqGLkWjJ3v7rZgugbV4W0+/VGDtlraxZYW9dXAZNvVb i5a5+JptRYND27b023y23xf5bFfx3a4/Pfvg0J8cydljTVc/8nDT7+UmUWeu97hfQrVxXeMW SciTrctMfHBPsiUZ87hbLcuMLfYMoVx8Jm6Ln7RhxzJri+UvqyVdcbb81j3L0UGLLPdycxGW peJyW9Hyhm3bljlbDKcgttieZWqLLWuq8z1uS+tYi8f9PIjArX47hkI8I3Ig3S2WS5Q5GbMt k8FYGy5YtLd/OZrcJr3P48auuBhHcWvA4/Z63Jc97qZlQ7UHeFAi9qsgDMl0Kr1/9Spi4wTz BDfCnadL5GUyQvpIE4mQOuIjTuJA+NyAV+EyvAwX4Tycg7NwBhZgHo7AYRiDIUhBEvqhD3ph M/RAJyRgA7RADIJgFXz85/gRvox7jLvAfoE9xy6yLlbHvMecYh5gDjNTzDjTzWjoD+gb9Aw9 TufoQTpFe6mdltECkiY3yO/I++Q9cpX8irxFfkreJN8jl8mL5HnyHDlFUmSA9JM2WdZ6lNUj S2sjVlJCCoiZ5BIN4Qkry38N/gveg3fht3AV/g2uwM/gLfgBvAGvZTV7CZbhAup1GCZhXNYo IckfLVf4OTfnVJdqdJxzPjdH9Qn1Aj+vmFFOhcf5+flwzjpY+hrTsFjicgGVSIXDH+5GBZRH i0KUB54NgQAKyn6Swa5PKgjwVMhjXDXg8ltFq+jATIAF640YB9fByr54PYbIuoiTvMc1gR6C UTdBVZyiVuQ1RE0EwjkVghpnVBOFYBQ1CkYH5DCbSzhixIkx+UkwYDKKVmOlrVwRsoq8LRQI 00WXaCAl3339u2+QwkPfPPtbrsmUe701yfNJ1uZMrkMqgYPIGWX4Vj/EovW+0nyWNYJgrLJr QKnAxYR8QRDxxT6iLRKqiZJnmX0KrcBZiT9fw+wtUjE8DUgioHIoRKCR1IXrgow1VFtHGkmo 1ktt5byilBgNvK28MlTbiHKWUqMhlyoUVLPxiRNcmW6w8cbDSrJ/8Yy9dbgpeldVoLW/oWHr 0df3jn3jvq475r9W+msChONt59P1LUoNmTtx5Ujkrh6P3S50RFq6ulqGT7WHdz98Z9/Dh+4r eF7SKbV6lQ2iThthfTQS9pUoWHaNwHoNqNOaaodsmhxVVzMFlTBe3ghcCdlXACzf2R5WsHer tIQfD7rQipsyejVJitU6Kj+qiFmq5hKFSS7wCuzxkqyOcoMpGAjXSctQ6fASFymvjJDacLAu HKb/ZHdvbzlUHWzrXVvfN/3s0Ov3rF3fcV/v3anYxq6WU28f+P6iO1HlaHe3BFNP+O4oLNbr iwu2BE6odVr13jsnWfu4K+y084l1se7u2MDi+vbyvc1bFlPFuqj7wc39nwrsaV0a8bdXVLT7 G7ty1jsqqyY3bpy0V1ZrVXrBoOSfRkeGuxW9FRcDExRAJFprNlAdcXIKFa8leSSHCE61DLl8 rKkFY4FBy+ikOskns4KugCNFMvK83pvYCxptxmAGf8GQNUTCjEiYdR7R8Oqrrz759JNP/2rl Wfr93KNpJRcz5d1YTP5CwuEvksyS/kaSHl5ZlGRC78k/ht6KQgVsiDaaDJRXC1UE91SpyDAV Fi1TnkvVTo0sGZXqGsFYpLGwDGF1JoOgwJG9GhH3RaUkXZPLKxnPyugRkozBbMrKh1YJS/hc RwJmkmmjNeSHV//+5bmVn04HG9pPfHxmNthRGGaoii/T2StwbV774ue/+XXyDl13o5640u8K KqUkP8Oo1Ozvcs32a7/BKmJvYfUqct9ZaII1UZ9Hr2Xy7QJFRsp3KjUapb2qVoafnqXOMsFP zmicSm68rAEZovmDXRS2Bsx1YRQOV9XAS7hxeBlbOSItF2EloS0YkFCVGWLgWQl3Di/FljAZ uUJmf3bxLJ+Dttapg2otVXXd89m+/kdO1iZCdYmN/3rvPT86qlc5Vfqn0r9+dPpH39mSGAkG R9nv/d0Di7yoTudKGNv+yGzM1rFmy68+de7tzXdsGv/6PpUoqs68PV+1ta35qW8Pj7z0ZbQR xkPcy9wRMIMd2WNtAaFOTkOVrqI8dDjIxkYiOvVoKoY1E71gZIHqKvTUKgrjKitwhWRKpISn jgyDufx1QdEWsgZubSarmDENQdK8tY8qe956i0nc+OT6TT0bzl1PP/fWs88/+3V6LX2N8L2L vgq9vqIryb6QnAx8emvfJ4J3rfwOzcKVrHze4rZN3HHHlDWBTL2Q3s9PoY18yBDboTMa793c E9NzfSYh7OGRJBLEv76P5Z2OYsZksCm0qEBvT8LB+rcJZURPJnL8Bi5Gzm5rIfxEBBHL7Lhl O32tri5cEQywZpnliELiDGQBNI20SbBFtpXMFI5KapeZ4TbTmsyVmaFmSeEwQQapk5gH+xZG RsdIyWOfIaVj8Z0u10Dr4GHnlrreIYOYZ4ocHdg95bWO7tjznUvrx4bZZ9zR+udGW46NV8Vc rtia7fxcz5YD9YEKj0+bzsktTG4KzIQS1fS5R4lxaCSVGon+zUvDqReejG3zr23yOPNMprzd X90/tbSkd26ND5Oc8xfe3rF5R3Vn/a5/P3v6ys6NXlN+x8BAr6Mm+vATK/9CrefXVJb4ZU5Z fZ89yHwOVzYice+akNdhZ3i3X8A/FSiLid9XJmhxAxQIa0gVOhoSIsVkPrdKz3ntdBZ4wh9Z V44r2nQb92YZN5whXNGAq1NC5H0RIdJ6Sn11YXtI5miFSY97Q+rIrCg521Fb2/WTz7w8t6Z1 SVLsyJsVX0w+/h9zd3yFmOrrdg3FfH1NU7GBU431bRP0SkM83rA/0VNxKJYu9hlEk2iYt/Xs nJjvPV7Wqugdb4jklq8UhAfqf8iVWopG18V3SFonUOs3uCX03D703dUVhPU7VKyykKh9Zllb m6ZYmPFgLFdObEWFzLyymONpjRwSNDWhhhkvKcpIyeBcLyvmUNhESauQXcz4G8QUz5wxRDdU bX7Q6zrS0XJ3uUOvP9H+8njfufrhZH7BTjIRubD5nz/ujpQm6DulPvP6xjSNddbWhI9NTi6u PB+LoH7sFYdSndSuPOWrnXdOzzek/lbymt2r7zP/gLYLQiBaXVOYT1lWK7AeN9rNTLTuUlmT QkLJrBdEbkZVgaQVyoQ1koOUYfoHBCVRmNlkzKWSJhnHWNdIg5aMC6XPOVs2rR0qsxvKHPb9 Tc0j7TrW5FDpc6a/vGvsWxtaN1iiHt89x1vJhGRAemWtQxXy1y8O3flQsCZYvSmko4x5y3Rd JNb0lVR0ftrpXlv0Alovi0KMNz+H8W4thKIek0HH8BrBi0YBZcCJLkLjLkQaZlj0LycZJ+Fq yYzCouRp+JY6GUdvNZRSySQS0X5EE9RRn9m9ktoK5nhj0JS3crigc2b92F8WC4VmJWsuOvrE 1onXuts7S5u8G14c+3Ep01icqqsftpQaKeNpuJjKM+1YGArs7M2niqqNdwUb401fGWk8MlUd i6RTj/dXe0M+3/QmSaedeElwZyAPiqA6atHq8eSoJGyO7JyVqgI11SuAWVBSA0dKMshC5ydx ahZAGS7VZ3eS5DZ2njabtpPh+cOYfh7bMRBv3T6QpKdtGmVKvPEK8ibbu6u5eWB7LLZTwock QT+iPA/cURvcFiMQNiOEBqNToBkRxKwIH5Vg5+kCc5JMLh09euR4kl2QkXjtB5LzxHfciTHJ O8wFqIGt0U6OoeUGQUlURCTl/lxVjrKmqojxVxcy3nxqcJrkIECqmgQjhi0qEynCkfP6EiNX TWp4qlTQ+Spi5Gnw5k6TxMlI04QHBsQi8qwcuWSiFlHyL0jGNNMkN17NNycffaV0z/2dHcfc 3aeP33f8+EMqsXBN41Sff2k8ZI0duw/XrEKtTPFq+wvxkfri6rZHVqRgIEn1zVtKXb6+A43p 9UmW8smbDNmL2DSBF2zR4hI/h6B0ukU5LJhRm4GbsaFLof7sgUEvcdvNra/IsEJl5U2/cfuG w3Hh2IbWh37zm9PeJkt7S+Ly3onLiZZ2RqMtKzsQa9pvLdNqy8qnm2P7LWVa8uLI15oXFlzT 98aSlxqikUi04dLhkfs9/kDA7/vYyORDgWBNIOA7I/Nbej/7BlpFi6eErmjcRhiV4CSVhPHY VRqliah8BiSHUlzSImJ2Figkj4/nMYWxQjXjzQOugChIRSm1k3mzpSRzXpCxITEfuY350EvK vOe43SKIHOOtYJtnjhQna3se9CP31e+xdh9A5jtz96mLff0XT9Ug9zU+tvXNj8XiCXrFs6Yl kuZjnXX+8a9eWrkei+xMI9zYJFtWIabEYX9oPnppNxIfwVOfdBa+ADbYGI1b9SJDeaEQo1zq N6F2OXgIUhHeqZTRpiFKwWhUGVRKPNRhXDdXiicTQw41GekCn4ekbs/6rYx2tXXouW9T5iby wtkDgkSDDzhn7n/m9OljR08hez/yjc4X64abO2sCXfS9tRVvfzH9aRlJNwH26juVXaH6lvhN vu5GJK2HRLR5Xb2eYaoFNZ6ClYQJ8iq1MkJqcHmrfW5Z8havrVzJmcoEB2mupmfyTCIXwvh3 qQzwBNd+W/wSlpnNLIHupptFyviA3MtIBooovEMhe2axNsOPGKqgBSt9pLKyO1Sz7mTf/k0b lEa1KKox2Dw83thvyy/Z3vVo98ktgWDy+RODO5VmlS4vx2lmI7+8pzE01P3pe56kZyJ7y6sK nXZWm/OWRtSrtu3KNecESryNeYW5PX/R1XqosiMe4XUqUqQRje87KzvqN2mKTR/D9XDhemg4 A5RBXdRrIEIeBmsEKUqhYpR57gJ5g/EC+YRRh4dJ7OHIvLo0h6dWGY7SQUEfksLqIJ6iraJ0 WAiEI3g6kGM2ifVF146TG1LEmX6zrCgSakk2TjQ/1u92M57EgRfvSs8lVy7f6XQMts/Oxmda m6yjdXSdxGi43yVGq8LzgMeGm0brKxd0yJlWoUglqDBIFgQrKSTlZL7IaOSUZF5AMWn1TRxh eGCXljhDoQaz5JBQyA8a8EzDhWReIJ27kazInEHnDrij01Fz/i5yqMA0mD5b/WDfwIOeyP0k LeFIXOmfM2pFtm22lfx1pgEcgiqp+ULHVr30MdCOvyS8TnaQL1ElPUp/wpxgKTvMvs7Vcge5 L/H5/D7+54qI4rJSqxwHwh4np4HD2O0CuRsgncrc2c/CCNUhfai5Dz46Mlr4UIoCvLvKXUin ABTr6KFjH/SQTH7cd1W3K6/hP4FRfvhZWH1g5eNCt2IdFnnkqVvP4XzdAKo67N8vdMsz3Z7M fLaJvQAXMR/EnMIc49+HXv5xWOAT0MNfhQV2DmJcCSRYM3SzdRCjS7BTyth2pzRe6mNew/wF 6GbOgktql6c2wz5g5VdQ0KBs0pflNMxjm9SrhVduyVSW1VS68lgj2acUaINMmQEx80VXEhjH RLJlDnvi2TLGNNCWLSswd2IfYQUsNd4qE5TkZLZMIRfOZcsMlMPj2TKLY17JltGi8P1smcf2 H2fL0vy/jE9Nz+3bMzp2wFJTXx+ybN6ze2xw35Al4bV07Nm7eyzzsXnz8OjBicF9bVN7D3RN 7R0d3rc20565+r3+lsw36ThMwTTM4ZrtgVEYgwNgwQigHn8hLG3G1t3YOoj9Q1hPoOe0QAe2 7pXbh/Cagll8wo9jh3GGgzAhj27DeffibF3yfRT79sHaPxh/e9mP8/qh5bYWyHxj/x8Tza66 ITOWFGLm4f+eKMNyt9d5eZ1vJaXwofEqtQZy5P9oYECmFXWgNxhN8OeTWGqR8Soj/t3V1VW8 gnQFkt0Z/9+JokVY4P5kP5/dK388KUH4X+ZXgRqk/97kZOu58jVP3t8iIPPhedAARvhzssqf Ufpv9p3ZO2VuZHN0cmVhbQplbmRvYmoKOSAwIG9iaiA8PAovVHlwZSAvRm9udAovU3VidHlw ZSAvVHJ1ZVR5cGUKL0ZpcnN0Q2hhciA0OQovTGFzdENoYXIgMTE3Ci9XaWR0aHMgMzcgMCBS Ci9CYXNlRm9udCAvVFRkY2J4MTAKL0ZvbnREZXNjcmlwdG9yIDcgMCBSCj4+IGVuZG9iago3 IDAgb2JqIDw8Ci9Bc2NlbnQgNzAwCi9DYXBIZWlnaHQgNjk5Ci9EZXNjZW50IC0yMTEKL0Zv bnROYW1lIC9UVGRjYngxMAovSXRhbGljQW5nbGUgMAovU3RlbVYgMAovWEhlaWdodCA0NTYK L0ZvbnRCQm94IFstNTcgLTMwOCAxMTYzIDkwM10KL0ZsYWdzIDQKL0ZvbnRGaWxlMiA4IDAg Ugo+PiBlbmRvYmoKMzcgMCBvYmoKWzU3NSA1NzUgNTc1IDU3NSAwIDAgMCAwIDAgMCAwIDAg MCAwIDAgMCA4NjkgMCA4MzAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgNzg2IDg2NCAwIDAg MCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgMCA1NTkgNjM5IDUxMSAwIDUyNyAwIDAgMCAzMTkg MCAwIDMxOSAwIDYzOSA1NzUgNjM5IDAgNDc0IDQ1NCA0NDcgNjM5IF0KZW5kb2JqCjUgMCBv YmogPDwKL0xlbmd0aDEgMTY3OTYKL0xlbmd0aCAxMTMzNiAgICAgCi9GaWx0ZXIgL0ZsYXRl RGVjb2RlCj4+CnN0cmVhbQp42rV7CXxb1ZnvOXeTriTL2lfLlixLsrUv1mZbi+U1duw4juM4 jpPYSWxnIfGajUATQhYIhLAkYZ1AKQ00TIOhFJgOXWjLVobptDx+nb55HV6HYTo82k6nM/Dr kFh537mSHSdA2/f7vVEi33PPvbr327//9517EUYIydEhRKPhFasCYXXdfa0I1b0Ms/09Tcv7 Tn9z6k0Y/xq+sc07RybFK6ovIJQeRoiu27xn1rrhtpHjCFGXEcKKscnxnZTy+yMwroLv34/f sH+s51XuHxFiahGKXt46OrJl7N0OJ0L1GnK9rTBRMs5GYH857Fdt3Tm7zzWuOwL7k+R6N0xs HnkBPfUS0BOE4yd2juybxOsILQ2HYN+6a2Tn6PHbPtgH+4/B+ecmJ2ZmgQ/4pLPk+OT06GTP 5p+dhvvLYL8aAa/kh9QF+jFyFnseoS2bp0Pk4gufK/CJX7mSL+xdukQulNVpNWqVUlEqL5FJ JbxYxLEMTWHkfQ4u12Q9semEPTfW7/M+50aoqW/fHPIM6JNrhU2DsJmTuNfO4SbFdVM+rwf+ z1Fun7fwB2+aQ/bcs5h25Kyt1q0jW+ZoB2ytc0yTfcS65dls9kTfgA0+zyJb2dqeATJgHNbW rXMiR8vIHFo5YDPPsU2JteaRLWtXDdgD9v8yB+wfwr09X3z1kS1feCbryGU+/MyPM/8lTClV +qTP+wqwDXJtshIiT7TM0U0j24abh3M+b4t1jrbnnsMU27Snyd50wg4H5hh7zjAnJbzC1Qwt 2+ayI1aB7rlGM4zJcK7VPNe+du0ca8/BJbA9t22OsufmZG6Dz2tt2drs834LKRF75Xs5IOI5 JYvI1mp9lqJP5OxzeCTXCgLL9g3MZYfXkvv5vHCoSdkE5IaQz+v3eV/1eTNzGrcPcUgMtu9A EbQ635+fufIOWPVO+jF2J3sev4y34g6wtZ+jn6GfoLfRK+giego9js6hW9B+tBftRFvQBjSE BlEv6kbNKIcyKIWSKI6icL0QCiAf8iAXsokui37N/oC9yD7C7mf3srvZHuY95h8YjmHoy/Sv 6A/op+g7aD2toi5Tv6d+ST1FbaJ6KRtVQRnxe/gF/AS+HR/Au/EQdqD3BWreEeh5HX0HvYy+ hV5Az6Pn0Bz6S3QB6HsCPQYUPoIeQg+g+9ApdBLdjo4DxQcFmrejrWgM6B4CmjuK9C7QakMV 2XJxQOTirSUqJjAJtn5YMSka5zZ517OTk5GSZnTyJbrhqMXjQRQJGiz8A28TocqsOUxxiGPC iEciijlGw6FjIow4ii+lPSHkCdqUNqULvhgxyHY5x6JLyMZ851IOLOfNPKLOcwjur8uWWjUK ilLt5U2IxXb4rcfvD2LwPpEcS7BWY8I6Pdnj7JUupwRHa004HouEqccZbvi21VVVq28b5hiO lzNnf7hhww/PMnKe/VTM8dzg+/lL+X+/80T+3/KX3h+ECTHHzOImfONbb+EbcdMsA3TMAB09 QIed0CHHVhXQITICHY4iHepIWK8zARUSLOI8GCiI1sbTlCpaS/mxsBM7ytGSwq0lNLkHN3xi pcOx8gShikPMbP7b+WNvvZU/lv/2LMMJhH1wGZecOYtLLn0AZEkYuD9Gc/kZfA40XY6sWQPC YhXFYxFmlCIdSx2VljEstgpizXiCDpAE3BlkAGKRYyKYAC7QFa114omKaDiZiLSmqrvijFTC ABW1bQ7scOQ83uwtfZ13HaqFOZoXicOzR3JSObm7Ht9MRfF/gmb1WYWFBi1iphtRIAi2qExs i9qo6Pw/UeX45r3wi4vws+NArxQZskpaxYMRYGYtsMLikuJP4kUSCbkXsz5/Ou33ZXcWNhBh KdRz5R36L9l+VIHqkCdr57E9xNv9ELbFoYBGsCoej8vjmB13m8GuGgrX9QTVaSoSLqf05TRc HKwEru/yA+9pHHP4YQ/mKK2mHOvLcSQci6fBZpzOl6s272sKbVi1TFfl1Yo9ueDFzZu/HvaG SkXGGqqhbKa9daZcq2ZZjQbG7btN5aWsmnm8boXH4G9zJY7c+GDrV/LbVm7cuLL96a/tar3j 7gNVzf19raGHH5+68dmnou0bNrZ1z3976vGHQTYvgIsMAl8mVJO1YazDLqNBCRItKcVSXmtU SGkVwrsYHWZxGbEzjx/kS0Sl1ICpR7ROe6UoGiZ0Kzl7NEx5jBKZQe258NUnz3/1gkejnzj3 B7a/RHKpQ2dnSjIclylhbXomG0vDvcfg3mfh3kHUlE0GjHpag3iNywkyZcGbkEnEK0GuFZIA VpTxbswx1DCr4NlKHNRL6MEyMc1RYUITsTQwNWs8Fo/QNjAunKajtQXpXlUrkTmoAvxDTolE +F3cKzEYa+vmJQyOdW8ciGX29298yKqRejJZt7dxbG70S69P9k78gvv1OZn50fwjHkaMB/un bn+wa+ihwb0bmGQ4nEiGRufGxl+cWv/lm37IfQAc9YOVnAeOOlFztj7iK5OLWTbGs14t8BSr cQp20t5KISk/ZKtHbBneaEQM1xwR00NSJeYGgm6IUl0FnoAlXOsqcHEtDyTEcCKdnvAimJQf Rxftt2BHsSj5KQSBFK6NELOKUXMBY4d78KyjvjnlqcmOnV/34MTMo4duWpfKHcw8unr8+7ub A76mjlTNM9vD/eX19ZaB0J0qlaM7tynEpHbGy6taGmojgXjC33e4pWfVnkOveHTd0Tu7n5xK 1vXcPxJrSEVj9e39Ye/AI48M+CJqTgOOO3I/eE4StPwDdicyQ7yIZj1lRkqPXQwnASOzlCpx Ca/FsnIjR6tKMI1V2IJ3iPWWQgjxZCCuCfamVUa0dq1micXFUuDmOEYrMfWWRv/22zbr0RN3 3nbqqN02/0vqgGQor2R36i+nMm+FpQW7k4bfytBvl1x2Un+YFxPrA12dFXTVlK2PRnx0c4Iv oSgDNoh0Sm6ZkWr2myWMOCGojWZUCmqTNY3ZII66KZaiDhqRlqO6Fv28Nh4mDiH4sKAKcGzI AkqiF/3C7lJFVbqWuLygZyFqR3+soRnqxOmOk+0UQ4nkjOfAXV1Da7ae67kzPnGkZmwiNfq1 zQ01nmRDtaa04vz2+h12h0leYeoLvjTIyXgDm1EzGOeslRjLxCVMajBozk9sb8seefHbx2tS FePfGA3GoiF/jBs73BCJ3bV+/Ja6SPpMiZ5kmCs/od9jt0Oms2fLRODxCmySAZeYMdrkLK3a qCtbTHugFauIE9nVESEO2IBBLKcXrBTSnprqzd8jMRkjjRGHufETfAuTfz+esogTTmey9dC6 7zK3S8oew09qLazMxYjzu7pffOfV055orfcBXIUVh7CMRF0hctMzELn1JIsjFSdBYpVKJjiS FpfitTwtYbFxIfBfzTUcvWSMzRDFi0Fd2FIPZb1k6M3Ojy0cEO4Hme0O4X5OEg8L95NhjUoO GY5iykQ2JU3z+KhOxrK4eiHJLb2tejHhER/VXs13OLpIhdboJllvckUHIxVDyuOaly0h6GY9 SX7vTr7dTTHgQxJW3PHyGNA2A37UCxWBFDyqIRsxaUppJccr/TaSgzCnQrwCSCzFMp4fgxjK 1hLhYE6DR/QWzNH1i1HFtpRcMD3n50QYElpIarJgTSGIpAHI4FeLDHxDbk4eHLjbVpuKOZ2J 4YfX/NV4uL4u0jub1ZSoPAPp5IJMmVPNXeaGjsaA3xOJeHoPtbTbplLpnfbEyQsD4yrV+Lq5 x9IZ4C0Hufhb7Msw8qBQttqFRSSru8qUFKOucajpUl7rLpVAWUOqGknCoQHj9BWU7ib5nsax KG2juWsjRBzH6jF4H16SqfA/4Lc/wCsqZvKijUyFr3vfnvHRwWSdDs+JTAqpmsavKuz2fUcO n3lgE6WiHrn0hMSHO/K/5VVOk4uEEZsty3ygK+E/fb8YVwCHdAD1UXYv+E0ENaGubAuE8gTC divF5zDnEssosRoHEnbM1WCrqwLqM5qxiiiVEzdU8Uq83hjArBpvkDcwXByPV9Fijm4Rcq3H HYRIrqx12qO2sK6oFE4koAVhaLcBRzYl4dq2GEqcmNMqbeFiDrDbKj+1t56kfjPwZGPDsrb0 hom22+Kp9rb0rS9f/vlJ6ub8jErRjrn0Lo/PV12zMq9W4fitA0PLrSOh/Kfty2UhZjxkSEfC /ltaW/b7G7Vc1BM81t51W2QiQx2efyMkKwl4A1Op5j2BEeoGWWg+P9h/cLbP6RfiyTsssdgU lLu1CZ+KNjqIRWKjSyKTiR3VEcGNNRINR7kTfAAflrvFbDmihhL1Yo7KXE3tOFZFomhcUCcg DwHd+mlit5xSAdgqEiZ4Eo4SkAlhlnI5YV+Fm7dg7tTEd9p30yUcL5OLdCqaadv32PrZv2k7 frxqcGxXfP2Tt7SqZNq7Prmn7cjR2eiWnTuZt8Zn169+SiLj5/+Rl6lHX7qzZ2zoYv61/OsX fZ1h87Ljcxs0mnv+cK+5bkV0O1adPo1VYLftpFoHGzCDDbizVWHMu6Sg9ypschlB4RRjklCq 9UoPYkcrGACIUUHBQQE4CT4oaFinElAy8EAUC2oEzQJXsQJcLmT0T5QajdJ+B/VvN/9r3+BQ zaq6/n+5eV55BxYrFOqWH2/7btYXVvgbU9+94ccty7lKLTOureRk6YlNW/PzJ9c881hm4/YM 81tZ6NKDoLqhVfX3Ptp3/z8PrBkBfTULdjyDEiibTVpKKE5XgSinhwIbLVixBNdil86pATP1 uGKCHcdry3kRq3LhTVoaYFudwNYSs1XbofYQ6AZuiukuorQXmQEQCexBbElTAnQL/yeYKnMw GZH7EvFzN/+4xf7w8k2WCnkkUlpu3rL8YXvLe6unprzuaLk80n9p9CQxz+GwIZuP7D3nDxsM 4by0tKxiXfv8v7Svt5gV8++GjOKqase+o2vnf58pcMi8BxzWomQ2UC6HIkjPq4rMufQu0AsT tIaxn9dG/BWEr2q8SUf4ihWwr5+obAlr1B9l7WrUEdjalIxI1V5H7MF9r7TYz3RuKCsvJXyZ hjvPGNS+i88888CDSxhaPb61gnA0/7vSMsv6ZfPvL9tIOPq3BRxd9K/3wL86UW+2rSmXTaeq ZCJPFACVAfRTBj7G4A6KilYXgLRTgvi0gaGdMr4Jt+E6cuyg1WlhgypqSFYqLqAZ4nIZIUu4 nK6lzhT/E04IaSONIVOAaAAGxbUAZ+AsiNkpRryyyv/uTQsOtuWPOOXtnT+d7rtvbf+Fp5+Z 3b3BqGNYJpdatX6Jr237Au/8wYUN48kdzXNBUxHt5W8QM+KiVb8EntmAerJZlypIMkvCVRer oR28tkxmkhlkUhkvE8nqHRYRo4pFvG6DVi4Tq5UQkjbZ1BQbwpuUEJSptGAEGeEjANMFU7jK r01Z61qcIPwvSUdRZSFSL7GJN8Y3/+iuSju1yW4/+aPR0TdOtraePHv//Q+ctNvnz9kr77r/ 7INnTy7EYIZJy4AtMf5DWCojewbixMYiu1Rg/icLKQm4jgOceQq4LkeOrNlUhg28tsClxSAD LjepaLSAsQkz17Mix9cWeL+rWnb6p1tHf3q6SqVWq6pO/8VDj5w7vZQwiDSXHtTYuQXbJFJ/ D+7fDLE/lI17lYw6l/bTbl5rkZllxqLIm9K1Abe93KiRiaSqpCBmFrcWqSKEuT9HxpgktiUz Ih6T3LhAaqH4waIvkvXRCpG6ZW7NZF1XZ1t7/gz+QcxbXh88sO+uk95Yqbgif/Kjzxf8/AZq q8VWIvWRfastlz8/P9qvL+UIw1KdVGH0Uc5f5m9dooNuhEQMyIDgtZDfZTOyrDroNdN6XlvF S2SULKSXIkblLS9jGdZWzW9T2hB7S3UMwlD9gqmBBNTgcHqeYBnb52iGyAeTagJfZ2wgtO7f 3Gbi8Kv4Vd7Ygx132Inu7Hfcd+/dZ0AQeUqq1knz4m3b9+3bt3tbNNJ6B7tXf2mCuUd/nTrB upjnP/2fej2ru3S6sXEBBgVCBrTEv8Iolq1xK3laLcUiXuuXVckqZVaZHrwo5HaolXIZg/k6 q4hmcW0BORc1DIQTKKMFSKqPkq6VkkA2kvaAiXgsEiXcaUnVq8ROu42MsL7MFLphz/59+24I Rzkx3SuSyfiSWGT7Yyt6HtsRic47WHn+2MMBivVcOmupVdgEmq2NTCMnu2SSlEiYR5usVkGR 1qbLB5mEJC2jCMaGiPoDiKhGFCI1UIWXBXDt8RQaPOtkJsQOVSFQT6RYcKgV8UJToZC5qWKD x0k5FMXKnVMuNHcIglbNvLbh0u1DF8/khsdN4fbIxsvHjv/Xpk1bmJtvTn1zbPTFhoP6ZCD9 7Lamwzuqm5PUCzj2i627Y8Nt2/Mfne6888jexPj01NYdWP4crpkb2LFj4Ov5/7HhG19r3TAV XZ/ahpkCDmEugzacYHWZbFSEeQ+fKIIRt8dAMlxlFbbyWgfgEZfVQgdN/DoIiqwbbzYxeMHy PP7PgpNyqtjCsl1jgEW0WSxjqSUI5Z31z7bWZKuWN7Y+uz7/PDG9+8/ef5YgFWy60F9dZnLb egeizy2FKfGEbfXmaNdtoUQ8zyzYH4Er83/TXz27vHfSZXMoSKXWDpr6uRBfVqG12R65iOKX NTmpRDnF9xbZ1eHAsgTmsjiXpsoDXt4AcCWH07y2CVhP11NdSf5wTSVidViO1ym7WK4djyYZ AN+rCyIgwec6CSwC8ApcThGVQiElpENSO/kpyH5EAEsjDxawgRCSnEtEc/PWN7pa16xp7ds7 dqe8TG6J+ZxpWwkvksg2r15+aCTibFlf+6m98s7TZ84IAnvvxfXd3etf3LtUVrG4/8Tyzlv8 cUvFbSPPD2xtqB2I8iUyycZTzuW7Wp6vHWx1zi9UKYII3+9rPHeusa+AHOjLYOfLUHs20xKt KNeoOH0Dz2IR1tdIATh4cA1uqA7wKrD7Fk9FuYjVJHg7zuLDpRoNa8YIH06EVRzVuVhVqmsF V7iKDwR4IHQ1tGAu7KL1VLoKfQ5lbUFyxDeIfABnOPzYOTMdP939d3v6RDyAA61ERfPvfPOG nu7pASxdd0dXXTJxb/fHd/aKeaENIJv8uzu293Rb4t2+/CcDt3Uxxi/vjadPHQV4kPeoFG98 1LN8auheXaNh/9c39T2Yrs+dOi5W4Xe1Mv7IK6P9a/eFN3QmpVnd1FdAJqSK28/uRF4Uz/rN YEMqsCEPduNq7HA5eSlIwo2dvFYjUUg8TrUCkqiVgSTqL6Qr0oEUrEXQM+FUZFcvgb8CGlri NlSftlylbBvq3e1xWaWRGuY9V0Ra4fTuXTnUqlJatKHXXvnh9159fTljMzCDoRKZxeN84gB+ Ox858ITTY5GVhC6d19vYxVzjAn94F/zBgfxZq3JpjkW8UadRKuQyUaycNHVcV4NvJuMPspA5 QE1X8yv/mfwK6YX6xMTnzdrMzC23nL5vuMaC78EHmDJv1+7pzWN9ybgxH+JNPh0rgmzJcWmV zHPpF/SHUpW9zEX2rZWNlyI6oNJx5R2xQsBizdlwXZWaVsd5v8wrc8vKZQSKlQg0Y97tclTZ bRUmvUatkpfUBnigO72UbvJxB1mS9pxLaQXqXddPXM8MMUweIt3jnJHLb1CWVWz/0pdO3bvZ U447cUVAYSH7d9+9yVOBl+NkZXz1Y703Z1bU+o35aV4vMnD5p/JPSQmvBIURZkvclxE9fVdI VtiV18Dua1ngnKSX8hbgXMe4Lv2c8G8EG3tfyJXxbI1X5ceAhwqcS2RiGScLuCtM2lJOpJJi FR51YLaQK69izqBDea22NPrP0ZfymjYIdT7fa12258jh+x8YtCs4fAD3K+SWjuk9m0cHGlPe /G4An/c9cOYsgM7b8XnqlaBR4ENsLOHnz1Ahs8GgLGjRlpvfnx9c7KdCFO6/8iH9O/o+QJvN qDWbiuKwlw8HJBIkpqEGCXpTGs6qphJ+O6+A4GvhDxt1iE1hK54s9dEcwlGO2pFzKjm6tdhL g1r/ujYUQTXhWArgnaaQXIW/EHch3MSjV9tXaqH6IH4H4QZ3kzbj2pPd944H6lwOu7PW3xVo HKBtd605+uam7jU/mrKWp9e07rUH6gZzu7LPzySbt1NzpEe1bKy2z7mzPh8N5HKBW28N9ga3 tK7csuuGFcerfU56MDRka/N7A/N1TbO5/8W3VM02tY2SHveVD5ka9iRSIA+KZN02TPurJIxY j6V+jdCWs8qM/Li7FLEV2GrQ01NiI8tRvqJFA9MkUELEUC4EQ3WBV5G9UE84lEJTWUhAdMqx PVU/ttJYdXDFbas77JmnB1ee8HVEzKYd2FV7sm/gyw2ZLPV+IlPTHcprV4zs2fL6G/P/nm3v bT7IvN6pVaV187uzbe2NuQtD3xPwG+iwF3QYQOGs26eHG7BsKc96oVwUl3rMAvTRYw1NDbuR kh2T2ABDhBba3rg2FhEUQCyxuPQQAaQm9L7JOhfZFxZdiFapg3XOjvvrtjlbuly76nacMtKl SoVk4+k1zx91pi1N6cMv9ty7LulwJKmXPeG2bHzHa3+1K5q5cURB0aUr9jfdVr1rNrXxmdwd weE2bzgMcgfamSjI3QYVfVM2ocZeB89hr58G4euwFjv8RiGFaSAv8lgb5jVYBplNhCdLtHp2 LFwBvMSWFL3qgpgXhF6I12y0uGxHsr26wKSyuHZHjO2t7JMbOo/4TOat2CVRmLfn3wmd6l57 Lp1uan3SubNuzV0O+6GVTZPu6gojjRrbm+vxtwU1XLJXBaTqjP62zLK2xtxT69dfmP/beGZ0 Ff6wYyBVV3/f+uG7UMG2BP1UoSCKZn0qBcVIeE+gEtQTdDCIk3i0wgIFhRV4WuXg2SAeZ8sU hbWxgpIgy9ptmuLyBORn0I+walRgqbBgoRZ2BVWKqN8P6+efKnevPrb5nIFWS9Slrx2s37Ot ujGXTtU92tdziqfKKsvWRuo22+Ii2j71ff2Ov0jvuWlEQav3nLFl3M1Pb1z/tabssnj+qa4j EV+oMTp4L1gaJCrq54A6/KglW+fEVqVRhsU+wBIKl0oiEVMML1QO1VAo2CneKUL0Zosdc5CH 1diHpymFiKOCimJXkKxA2JcGOqgW7NFItDa+sMxX7MTHyWHOgiNU36+stqNHjszMbo8+8cRG 776Ub3Vt1OGIVpnN7s5Y4rd0ZL6/GN2MIepC9uMjQ96wvdlXEwjU2L2i0uUrZj8mTxQk8/30 e/RjKAvYcxzQ58qMnlrRAiUPCXzg7St8LNifFftayhVcHXaUDfBYNOwLi5CIZUqwVk3t844q 2T69iN5p7MBcDA9lUnWVVm5nI2I5etvVNSXBe/TFBaUCP/FKka7Q+610EWQldLiL7XoCroR1 /jhRbNwZJ+JZgOYFN8ULYGTJoQO13rEj+4c9e1M1yUqj3mG7uaev1r/xxMyArS1xftfjzbH6 zkzTl9e8eOREfzDq9qc2PFgTHOiPGJRKQ2yVa5mhBn8S3bfx/PJIdNPofXdv9cTd1FB60jXQ 4Y+k3GUVVnOVprt/qm7S05/1RhMry7yV+9vr11c5HB1dE+sbK0stw/GNnWtvvCnR2mrrT2XW WFsdtsS8yKKIRDtWp3O5su6a6g4Sq3JQ3TSDvydRfTYowU6/HASNmVpTLFhGG3itWGbAFjyp qtSAF8ShThOL6EkP1hbKGsFsSNum4N8ZSC5kdZ4UlpprC2pyvFKotMk8eEbBwn5pMm//yq2m /qns3S2tJvOO3TMz09vNhsrKzhUH2+FYdb87Mzu1PXo7kyYeLpPo5jsa1kZaWk/mDSFZIW3K QvhfD8Ui/feuym8ncxiKroypsAaCP2WHkBaimTtr1anKoPbEjFKNS3ktL6kopYwaMT3N05rC GlzRCRbIJeHqKhPqJWs8r5sM5i3YqjcOT9wwsWv7oDcrrDUtl4WoBJCZ1c2jInFM8+JiGCbr toxKkHUy648YeMyooyEpDQ5ql9gkFRK9LBEL1VRVmEUqEdbJKBrP6JVaFtcvABaonxy0SKv5 jGSBVEcsuhTGlGKop5aooJY6/+at0R0Tk5PTO0xmIbPhCh+u0s3Mbgg0NTd134sZib1KYV+x bv22GzL8K7h88l+Knb+SYEH6usuePzjDKmOW47K28uaD+btLgi5eK3QCggVboiTAnwZVZHUK FZaDlGVyrBZTBRnrFmR8jYgXaXzdBArH1bHt01OgctIqKtz108uFllzhDtwB+na0GmJEV4mY UkqoHHb7NRI56LUnWk6re5fHI1a2jNeKJJzEjp0Bvgzb8KTeqWEjOIGVeDKQ0nLL8SqOWkFM uQ1MmV6zmLDcS2zZgguPBhSjhVLzBQZ9XY/ojxj6Qecjo3e3hOXVD228547otumZmd2gDEdl Z/eh9mjR9o1yPewfXLbU+AtWBcb/3Y17WsIrB3fdQczfuGj9NyeL1n91qmGtLdJ/9+prPKLo 7SA/sh5RK8YSSOclgJMxEzFGA0IbjYhNj8vwpNKmYQM4BkmPiMlNPL5uAV1d4/ILYvp/cvmD roe2/AmPv8r08O4/z99JXv8B5HUDCpEniiS4wstXBMiCs9ejElCXBI+XGDE77ljScipkhWLT aWEVkKRyp9NRfIygAEsWHymKxZIt2dTXBpN7tjtzzc25Y3+V3DvmzrRS/1tfA/i1cdZZoxeb Lc7ppqbd9ho9fn54LpfrtDa5cxc3DT2ZO2ZtcGeent50ZzKeqk8m79y0bOOoJ5HOxDP3COgk P8PUgI4UKIaWZVNuTEv5IGRr2usFVekWAXC5FZeJtB6JEZfq+fFoKXn4RYw9NsqLp/ROC0cl ShcT+xfhYdfn6Szq0ArSoAC4+ipHo/VjPSZAxvv7BwgyPrPZbN7+QNfyRxbCSPSe1f2P7muk 3ovmajrDef3Kjftu+vrTeXm2/WD+XiizUwyTksmD9C8aFIqs7jgA5cNPE5wMWBM4jQKnlYA1 M9kohSGXesXApAHs0iQwSaKISKvFhjAPgQTXiDh6UmzQAtC0XgWaJGirBVCsj0eu54i7CjR1 BZxJFzsobx77dnLbyrIuv6nIkNm07asFlPnSip1Dp+wSb7Y3uH/VQDWNbvevjn/r0tRVdvDR swWI+fT823tu3IzHYutanHv3zq16CBUiPfUr4MxLsmoVdlspxl8hOFohLJZL5FIK8TyFR10I s+VqqgLP8AaoYPylS9BYbXypB10XZIiaqEUYQ8UDnLx7Z9xmv/XIkVkS4yVK0I5F27wlUz8Y htNpSUJHbX9mLH9xod4sCVIJR0AG+eqD5rH6tgSgskL1Qr0HXpRDK7Kt2ZDH7bCWSyU8beeT OIrpUAT40EAxbfcbhEogCmdopBKAYqAiFz7cADWNH4es5fQBURnNUc1LHr5yFkVfgYUnrbiF 1pWT9KXD+gXbdAZwAUfHFtpfxdUxJ7X/wM7pdHJj5+27zg4MPnJsdRsv5sQyhY2yHN8zlU1u 6T4xfjrgWfPIVHtSzIk0MrGUl1J7p456XMsT3QqT5uGtJzo71vWJ5CLcIWfkp78cCq5t3iS3 qG/dkBgrb0xGGVX+E5G8pNAB+pA6z/waWUj/Sot5BW/BGPMukYQWKzxGIaJwPD6sU2OWgyMs npSWlXBUhRBYiNepo2FhCdOmJQtVQvQQ6m4BeCpdq25MRbAt/55DE7Zk2yJb133pbjrim3hh PH88O//qroTJfHDowIEdT5w6RNWT6A21GYne1SiR9UFuw6X+SlABZmy8ScJLdFjD8zZsxJV4 0qTRsmI8yQOBlHshu5GnR53Ra0qxcLHB9pnaDId6SDb2yVT+7S07UwYyhlJsR/5nlUf6ls96 mk7gNwtQZyK9xhjpO9WD9xb2XYUa7MWVY+aCDBk5PQM0+7NWjVVSLimTyCRQm0gQX6bXqJWl JaJGmxhk5/5sF+2avpNIozd/biNNr5fk39WYzYNbRndNtFrkFPN/3hXLzcnVa7t6mv1hXf4j iUGvpz6MSGUEuahUSpF8/kVKoTZYNC4By1Sl53v0RY2z54DaHGrN1martbQ6xddKwpKgpEB5 iQQ8QYL5oKfa5bDbyozgfSqFvDQZJf205j+nnyaw4frTnNFCV00KMfG3IqMo/7HCbOgdGdl1 Q7tFSTNvzplLy4x96zds3Sbsv/Mdkb6qYd3Gju6Mt8bwicjEmETvvisiXEPIErguMbDy+bfw r5cFSpdMfIPSyNU2U0EM1hyIQU/1zl/UFzprjAVk0YLasslM3Kdh1I2pAO0h0atMYipqMVUX j9WGg9VVhd4iqFPMYqTLeHIgkLYlAiHYNU6TtLq011aBwd+d18JWfE2wqw1gOHHJDLa//B/K KrVteGTzxESNXsJduNVY6hke6e9f1dgYfQKLNTa1dWR4ZHLCrZaKvveQThIY2bxm1YrmzCsf 4ZSukhPYN5TKJdJfPGGLewTWK5rezp83WQqrI0ZeJRd/+n1TyiNA2wrBMvKDjAykEYOI7ov6 XSLyxLNPXSGxSMyCUYgkVrNOhlmVAotwhKHGgm41ixNXe40ENLHx65qrX2DTS2oNEczjF9rz P9ZYiIVPTrSblMy/viMu1detXtvdlfNH9PlHxr7aNRGuqgqHbK76W+bmqA/DsoKSJXpWBVq/ rDaUqwtKtmcuP/bJSFu131/t9FKOYraio8CbBUWyHpOEYtS8EjN+mQSLKaz2lglP6hlVylKS Bii+DZAkiysW0FO8Nk6gBUWa88JTmsVnMMnjJ5VJhmp/fe1UZxkT0VN/yZcw89/68dA+t8sm jtjn03q9iqO1qpaZVXi5XCrnLv9GV+04tjl/C6EqCvYXZc8LeajZadXTNSq+xm+CEi6IAxBt SYvDJKaYhLguJqE5XhvD9XbKaaGmgxktxGOVlFJjv56aoRT6QhYiSEHAQrbrQLvQ5Vh8zHRB /IWlFgu2Oa5rioC6qJfzT5BCZZpk2ieeaGvI/UXf9odIoy1ZYxmK37RuAGuxCyIonDK1NWrS kk7IYjlFOiEtu6uD3pkhbzjs9ftt9p6Vuz42XM4XMa4xRDV0qksAIy1IQY7q0apsaw3mMc/7 vSCFKojzahcW2jxxUzIqlOsBmV8WxXUMJaIoP54OqPSsAVdZKAeuQdSMphIkkVqC4gVZkP5G lDAHSAMXnqn4Iuno49esxA3i50Kh4JoDA9VB/IAp37dUInUgkT0t3QsioGdwj7YupaWUPfYt iUv/DKXyhesk0rzXFfQ6ZFDuyRaeeZijXfgJqCgtUFFCTKLUYhkmFbyOpY7IyGsRC6k2eNVp FiH81dci8Ex5HXk+1J8pW7+ak3Lk3YzIsmrml55Grzub3hSZ+dvlHMfRPM8FZ480mxGFo7SL fgnuXIbqIPv7KrADWxClosRSjriGUmbQadQs4qrNfKSSOpqoVrBHzdLiewoLPehrH5UnaKai +GqLVsP+kYM6Hd5iTUQTdZGWnLsrWXyPo7sf313eUJuoC7flanqSjETCsAyzYg3G9Du+rM+d u7V3+akvkdc7OE4k7vzWuFm+OH0wAtOMWMJ2/vWYRAl21UrXUK8Lkq0n/DmxFZcjighYCgIO KuLRSNhvLxfVSfmj7jrEasmLKGamYD5E5NfIXHeNyCNCD8xxnUbsS5/crcUSa10omQBWNvSJ ORYYZKiePoxPVsZD9XXhQKps/SpazHBigfGipr60euYnyxgyKRK3fG+8RMl7sj5PBhQ4/ZMO gW+eg3lzwW9YDvymGa1GY9lBr57yq3gThvDQgf3+CplcvDJCRTG10ChNi7MNQhRJ4po43xD1 0lvsNZhraaRW4r72iIWZjiMRV4wqPcWoQq9RFAuQPxlW9LHPb6Fqr+u40sVjcqiryYsXTlr2 Wb9qusW64aCvPxohrVaj2dsZq/v9R0IndhaCTchrTY0KPvkcF/9M3BG8zOsdsDf5SRe2ysuV Lu+d+jg7378Ye5gJh591ET/9+OMKiwoh5IB/29BF9BuAI9X4NP4pVU+9Sivph+j3GTOzjfkZ G2O/zyW4h0RS0bSYFR8Sf4/P8d+E1HiT5HfSUzKjbJ/srZLOkm/I4/Iz8v8oHS39kcKv+Ei5 VvlllVKVU/2TeqX6eU2N5qzmP7X7ta/ozLpp3R/0B/W/MSQMf22sM+41/sq0xkTe0MT0r/Dd CBwPP4SnEMr3F7bMV9AYBZRSUvbqi520Al33ySL0uyvs+Xw/QiIPJbwZWvzgwhf/ww7dxtKG jxEtvv636MrFPGlGe2DIgYUt/g6ul0NI/P6Vi/NvihXClZZ+mrniFG1Bb8J3hkqiOeoC0sP2 IoNQD3xfgO8YfPvhmyyOyXkXyZd+Hc1wCOXYC6iDJWMLamenUDObhGu3C8eaqfdRnIN9UR/q JvswnoFz2tn3UTszBePXUQfMu8RJ5ICtUbjXFErSJ1EzbJthm6SmUEa4/xTKwX1zTBLmYMtd gDHMFY8lYb6ZtsNvXbC1IJdwLIlccA8XbI1kTK4H81HyJfwyCEfh963MBfBMIo5mtBMxglgo JAN56oQ3jvcgRjiqQK8tyrGiqB3yl4U9XPyVCFUtCFZYJSyMGTinvjhm4Ui2OOZgr6k4FsG3 HY5hhicvDi+OMZKgY8UxBRn37uKYRlZ0rjhm4JzvFMdgheiN4piD+b8tjsn1/75pYnL/9Lbx rbPWUDIZtfZu27x1ZHqLtcNv7dy2a/NW4SXk3tHx3TeMTLdO7Jrtmtg1PjpdJ0wLf4L+YLPw onITmkCTaD+aBg8cR1vRLFATIj1kkKQV9cLsZpgdgeNbYL8D+eFvJ8zuEua3wN9pOD8IZ47C 73ejG4RzW+Gqu+BaXcJ2HI5NQ6a7evbVURCuGAR9LeyjwvvWf/RDFSWtKZyLTfDl0J/9oeBX DMshkZiXSGUIinF5qWAVSpVao9XpDUZkQuYySzmqsNpQpR0sofBxOF3VNW6P14f8gWAoHKmN xuKJZF19Q+rqxdNk1WwxIDTmrrt3E/pv/DCUVbBRwcp/d+XKFfiLyF+Ei97w//dDFXXFCLZK LF+MeLBUKSJvx5cIx+SotHg2CZdKpEJq0JsWPFKPDKQKBFkjZAYkZEHlgj9akQ22lchO3uC/ 7o4O5EQuqPZrkBt80ot8MOdHAbCbEAqjCKoFq42hOEqA/daBpzag1OdSnoZv5nOPZFEjyv0J vv9blfjf+fm/iYgNH2VuZHN0cmVhbQplbmRvYmoKNiAwIG9iaiA8PAovVHlwZSAvRm9udAov U3VidHlwZSAvVHJ1ZVR5cGUKL0ZpcnN0Q2hhciA0MAovTGFzdENoYXIgMTY0Ci9XaWR0aHMg MzggMCBSCi9CYXNlRm9udCAvVFRkY3IxMAovRm9udERlc2NyaXB0b3IgNCAwIFIKPj4gZW5k b2JqCjQgMCBvYmogPDwKL0FzY2VudCA3MDYKL0NhcEhlaWdodCA3MDYKL0Rlc2NlbnQgLTIx NwovRm9udE5hbWUgL1RUZGNyMTAKL0l0YWxpY0FuZ2xlIDAKL1N0ZW1WIDAKL1hIZWlnaHQg NDUzCi9Gb250QkJveCBbLTQwIC0yNTAgMTAwNyA4OTVdCi9GbGFncyA0Ci9Gb250RmlsZTIg NSAwIFIKPj4gZW5kb2JqCjM4IDAgb2JqClszODkgMzg5IDAgMCAyNzggMzMzIDI3OCAwIDUw MCA1MDAgNTAwIDUwMCA1MDAgNTAwIDAgNTAwIDAgMCAyNzggMjc4IDAgMCAwIDQ3MiAwIDc1 MCA3MDggNzIyIDc2NCA2ODAgNjUzIDc4NSA3NTAgMzYxIDAgNzc4IDAgOTE2IDc1MCA3Nzgg NjgwIDAgNzM2IDU1NSA3MjIgMCA3NTAgMTAyOCAwIDc1MCAwIDAgMCAwIDAgMCAwIDUwMCA1 NTUgNDQ0IDU1NSA0NDQgMzA1IDUwMCA1NTUgMjc4IDAgNTI4IDI3OCA4MzMgNTU1IDUwMCA1 NTUgNTI4IDM5MiAzOTQgMzg5IDU1NSA1MjggNzIyIDUyOCA1MjggNDQ0IDAgMCAwIDAgMCAw IDAgMCAwIDAgNjgwIDAgNzg1IDAgMCAwIDAgMCAwIDAgMCAwIDAgNTU1IDU1NSA3MjIgMCAw IDAgMCAwIDAgMCAwIDAgMCAwIDAgMCAwIDAgNjc1IF0KZW5kb2JqCjEzIDAgb2JqIDw8Ci9U eXBlIC9QYWdlcwovQ291bnQgMwovS2lkcyBbMiAwIFIgMTUgMCBSIDI0IDAgUl0KPj4gZW5k b2JqCjM5IDAgb2JqIDw8Ci9UeXBlIC9DYXRhbG9nCi9QYWdlcyAxMyAwIFIKL1BURVguRnVs bGJhbm5lciAoVGhpcyBpcyBwZGZUZVgsIFZlcnNpb24gMy4xNDE1OS0xLjAwYi1wcmV0ZXN0 LTIwMDIwMjExKQo+PiBlbmRvYmoKNDAgMCBvYmogPDwKL1Byb2R1Y2VyIChwZGZUZVgtMS4w Yi1wZGZjcnlwdCkKL0NyZWF0aW9uRGF0ZSAoRDoyMDA1MDIxNDEzNDQwMCkKPj4gZW5kb2Jq CnhyZWYKMCA0MQowMDAwMDAwMDAwIDY1NTM1IGYgCjAwMDAwMDI5ODkgMDAwMDAgbiAKMDAw MDAwMjg4NCAwMDAwMCBuIAowMDAwMDAwMDA5IDAwMDAwIG4gCjAwMDAwNDg1NTAgMDAwMDAg biAKMDAwMDAzNjk4NSAwMDAwMCBuIAowMDAwMDQ4NDE0IDAwMDAwIG4gCjAwMDAwMzY2MTEg MDAwMDAgbiAKMDAwMDAzMTUyNyAwMDAwMCBuIAowMDAwMDM2NDc0IDAwMDAwIG4gCjAwMDAw MzEwNTcgMDAwMDAgbiAKMDAwMDAyNDE2NyAwMDAwMCBuIAowMDAwMDMwOTE4IDAwMDAwIG4g CjAwMDAwNDkxMjYgMDAwMDAgbiAKMDAwMDAwNjMyOCAwMDAwMCBuIAowMDAwMDA2MjIwIDAw MDAwIG4gCjAwMDAwMDMwODAgMDAwMDAgbiAKMDAwMDAyMzgyOSAwMDAwMCBuIAowMDAwMDE5 Mjg2IDAwMDAwIG4gCjAwMDAwMjM2OTIgMDAwMDAgbiAKMDAwMDAxOTAxOCAwMDAwMCBuIAow MDAwMDE3MjQwIDAwMDAwIG4gCjAwMDAwMTg4ODIgMDAwMDAgbiAKMDAwMDAwOTc3MCAwMDAw MCBuIAowMDAwMDA5NjYyIDAwMDAwIG4gCjAwMDAwMDY0NDQgMDAwMDAgbiAKMDAwMDAxNjkx NyAwMDAwMCBuIAowMDAwMDEzODE3IDAwMDAwIG4gCjAwMDAwMTY3ODIgMDAwMDAgbiAKMDAw MDAxMzYxOCAwMDAwMCBuIAowMDAwMDA5ODk4IDAwMDAwIG4gCjAwMDAwMTM0ODEgMDAwMDAg biAKMDAwMDAxMzc5NCAwMDAwMCBuIAowMDAwMDE3MDkxIDAwMDAwIG4gCjAwMDAwMTkxOTEg MDAwMDAgbiAKMDAwMDAyNDAwOCAwMDAwMCBuIAowMDAwMDMxMjM4IDAwMDAwIG4gCjAwMDAw MzY3ODYgMDAwMDAgbiAKMDAwMDA0ODcyNCAwMDAwMCBuIAowMDAwMDQ5MTk4IDAwMDAwIG4g CjAwMDAwNDkzMjMgMDAwMDAgbiAKdHJhaWxlcgo8PAovU2l6ZSA0MQovUm9vdCAzOSAwIFIK L0luZm8gNDAgMCBSCj4+CnN0YXJ0eHJlZgo0OTQxMQolJUVPRgo= --Boundary_(ID_U3KS6C0RmmCtPus/Fz0qAA)-- From owner-reliable_computing [at] interval [dot] louisiana.edu Wed Feb 16 13:20:20 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GJKKD0000314 for ; Wed, 16 Feb 2005 13:20:20 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1GJKJjX000313 for reliable_computing-outgoing; Wed, 16 Feb 2005 13:20:19 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GJKART000309 for ; Wed, 16 Feb 2005 13:20:17 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j1GJK5w23518; Wed, 16 Feb 2005 12:20:05 -0700 (MST) Message-Id: <200502161920.j1GJK5w23518 [at] cs [dot] utep.edu> Date: Wed, 16 Feb 2005 12:20:06 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: Re: P.S. To: Bill.Walster [at] sun [dot] com Cc: reliable_computing [at] interval [dot] louisiana.edu MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: 7JSJAlWeE8YmoTtPehGxag== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Correction about monotonicity: If the function is monotonic in each of the variables, you do not need 2^n computations, only n+3 to find the range: first, you modify the 1st variable to see if the function if the function is increasing or decreasing over x1; then you change the 2nd variable only to see whether it is increasing or decreasing over x2, etc. After n such calls to f, you know where exactly f attains its maximum and where it attains its minimum, so you need 2 calls to f to find both upper endpoint and lower endpoint of the range. > Date: Wed, 16 Feb 2005 09:04:05 -0800 > From: "G. William Walster" > Subject: Re: P.S. > To: Bill.Walster [at] sun [dot] com > Cc: Ray Moore , Vladik Kreinovich , reliable_computing [at] interval [dot] louisiana.edu, chleb [at] math [dot] cas.cz > MIME-version: 1.0 > X-Accept-Language: en-us, en > User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7) Gecko/20040616 > > > > Sorry, everybody, but Ray. > > I intended this only for him, for now. > > The note I referred to is still a "work in progress". > > Nevertheless, I would appreciate any feedback regarding it. > > Thanks in advance, > > Bill > > > G. William Walster wrote: > > > > > P.S. > > > > Did you have any comments about the short note I sent > > regarding the use of Popper's ideas? > > > > Cheers, > > > > Bill > > > ------------- End Forwarded Message ------------- From owner-reliable_computing [at] interval [dot] louisiana.edu Wed Feb 16 13:41:32 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GJfVZB000436 for ; Wed, 16 Feb 2005 13:41:31 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1GJfVnN000435 for reliable_computing-outgoing; Wed, 16 Feb 2005 13:41:31 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GJfNVQ000431 for ; Wed, 16 Feb 2005 13:41:28 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j1GJfI723651; Wed, 16 Feb 2005 12:41:18 -0700 (MST) Message-Id: <200502161941.j1GJfI723651 [at] cs [dot] utep.edu> Date: Wed, 16 Feb 2005 12:41:18 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: NAFIPS'05 deadline extended To: reliable_computing [at] interval [dot] louisiana.edu, interval [at] cs [dot] utep.edu MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: GKBaMwaxMJpt8VvMrVzQSw== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear Friends, I have just learned that the deadline for submitting papers to Annual Conference of North American Information Processing Society NAFIPS'05, Ann Arbor, Michigan, June 22-25, 2005, has been extended to February 28, 2005. Because of the close connection between interval and fuzzy methods, this conference is always very much interested in interval-related talks. Just in case, the conference URL is http://www.nafips05.wayne.edu From owner-reliable_computing [at] interval [dot] louisiana.edu Wed Feb 16 15:31:51 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GLVpFO000611 for ; Wed, 16 Feb 2005 15:31:51 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1GLVp0D000610 for reliable_computing-outgoing; Wed, 16 Feb 2005 15:31:51 -0600 (CST) Received: from poczta.polsl.pl (castor.polsl.pl [157.158.3.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GLVab0000601 for ; Wed, 16 Feb 2005 15:31:47 -0600 (CST) Received: from pownuk ([193.25.187.101]) by poczta.polsl.pl with Microsoft SMTPSVC(6.0.3790.211); Wed, 16 Feb 2005 22:31:33 +0100 From: "Andrzej Pownuk" To: Subject: RE: a book of potential interest to interval researchers Date: Wed, 16 Feb 2005 22:31:37 +0100 Message-ID: X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook, Build 10.0.6626 In-Reply-To: <4213796C.3080008 [at] sun [dot] com> X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 Importance: Normal X-OriginalArrivalTime: 16 Feb 2005 21:31:33.0961 (UTC) FILETIME=[E3230B90:01C5146E] Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear interval researchers, I have general question related to the problem, what is interval arithmetic and what is not interval arithmetic. As far as I know interval arithmetic is based on the functions with interval parameters in the following form: z=f(x) e.g. [-1, 1] = sin([-10, 10]) or z=f(x,y) e.g. [2,5]=+([1,2],[1,3]) or [2,5]=[1,2]+[1,3] [1,6]=*([1,2],[1,3]) or [1,6]=[1,2]*[1,3] i.e. f=+,-,*,/. However, according to my knowledge in most of the books which are related to interval arithmetic there are information about systems of linear interval equations. In this case relation between input and output interval parameters is much more complicated. Y=hull{{x: Ax=B, A belong to [A], B belong to [B]} We have also united solution set, tolerable solution set, controllable solution set etc. My question is. **************************************************************************** Does theory of systems of linear interval equations belong to interval arithmetic or not? **************************************************************************** If more complicated relations on intervals (e.g. system of linear interval equations) also belong to interval arithmetic then maybe all functions on intervals are some extensions of interval arithmetic. I suspect that the authors of the presented book (i.e. Uncertain Input Data Problems and the Worst Scenario Method) work on some functions with intervals or set valued parameters (I am not sure, I do not know that book). Then from that point of view they are working on some extension of interval arithmetic. Regards, Andrzej Pownuk --------------------------------------------- Ph.D., research associate at: Chair of Theoretical Mechanics Faculty of Civil Engineering Silesian University of Technology ul. Krzywoustego 7 44-100  Gliwice, Poland Tel/fax: 0048 32 2371542 Mobile:  0048 606 550147 URL:    http://zeus.polsl.gliwice.pl/~pownuk E-mail: pownuk [at] zeus [dot] polsl.gliwice.pl --------------------------------------------- I agree. I have ordered it from Amazon.  Perhaps the authors don't know about intervals.  Many people still do not. The encouraging thing is that even point people are starting to realize the importance of input uncertainty.  This can only be good for us. :) Cheers, Bill P.S. If nothing else, this is a source of good examples for you to use in your talk. :) Ray Moore wrote: > It seems odd to me that there is no mention of interval computation in > the book. > > Ray Moore > > ----- Original Message ----- From: "Vladik Kreinovich" > To: > Cc: > Sent: Wednesday, February 16, 2005 12:27 AM > Subject: a book of potential interest to interval researchers > > >> Forwarding, from NA Digest >> ****************************************** >> From: Jan Chleboun >> Date: Wed, 9 Feb 2005 12:09:50 +0100 (CET) >> Subject: New Book on Uncertain Input Data >> >> Uncertain Input Data Problems and the Worst Scenario Method >> by Ivan Hlavacek, Jan Chleboun, and Ivo Babuska; >> North-Holland Series in Applied Mathematics and Mechanics, Volume 46, >> Elsevier, Amsterdam, December 2004. >> Hardback, xxvi+458 pages, ISBN: 0-444-51435-X >> >> This book deals with the impact of uncertainty in input data on the >> outputs of mathematical models. Uncertain inputs as scalars, tensors, >> functions, or domain boundaries are considered. In practical terms, >> material parameters or constitutive laws, for instance, are uncertain, >> and >> quantities as local temperature, local mechanical stress, or local >> displacement are monitored. The goal of the worst scenario method is to >> extremize the quantity over the set of uncertain input data. >> >> A general mathematical scheme of the worst scenario method, including >> approximation by finite element methods, is presented, and then >> applied to >> various state problems modeled by differential equations or variational >> inequalities: nonlinear heat flow, Timoshenko beam vibration and >> buckling, >> plate buckling, contact problems in elasticity and thermoelasticity with >> and without friction, and various models of plastic deformation, to list >> some of the topics. Dozens of examples, figures, and tables are included. >> >> More details can be found at >> http://books.elsevier.com/elsevier/?isbn=044451435X >> >> Jan Chleboun >> Mathematical Institute >> Academy of Sciences >> Zitna 25  ( TeX: \v{Z}itn\'a ) >> 115 67 Prague 1 >> Czech Republic >> e-mail: chleb [at] math [dot] cas.cz >> >> > From owner-reliable_computing [at] interval [dot] louisiana.edu Wed Feb 16 15:50:08 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GLo7hc000804 for ; Wed, 16 Feb 2005 15:50:07 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1GLo7Zo000803 for reliable_computing-outgoing; Wed, 16 Feb 2005 15:50:07 -0600 (CST) Received: from lakermmtao10.cox.net (lakermmtao10.cox.net [68.230.240.29]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GLnvlC000799 for ; Wed, 16 Feb 2005 15:50:03 -0600 (CST) Received: from Inspiron-8200 ([68.226.133.93]) by lakermmtao10.cox.net (InterMail vM.6.01.04.00 201-2131-117-20041022) with SMTP id <20050216214950.YRVI17761.lakermmtao10.cox.net@Inspiron-8200>; Wed, 16 Feb 2005 16:49:50 -0500 Message-Id: <2.2.32.20050216214944.0093e554 [at] pop [dot] louisiana.edu> X-Sender: rbk5287 [at] pop [dot] louisiana.edu X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Date: Wed, 16 Feb 2005 15:49:44 -0600 To: "Andrzej Pownuk" , From: "R. Baker Kearfott" Subject: RE: a book of potential interest to interval researchers Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by interval.louisiana.edu id j1GLo4lC000800 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk All, If the criterion to "belonging" to interval arithmetic is whether or not interval researchers work on it, then the result is a resounding "yes." Ask Jiri Rohn, G{\"o}tz Alefeld, or Sergey Shary, (among others) for example. I view "interval arithmetic" as striving to compute sharp bounds on ranges, in a way that takes account of machine arithmetic. This may, at times, go beyond use of the pre-programmed four basic "arithmetic" operations; I think the goal should take us wherever is appropriate, and we should not get too stuck on what the subject is called, except, of course, if it would otherwise lead to technical confusion, or for promotional purposes :-) Best regards, Baker At 10:31 PM 2/16/2005 +0100, Andrzej Pownuk wrote: >Dear interval researchers, > >I have general question related to the problem, >what is interval arithmetic and what is not interval arithmetic. > >As far as I know interval arithmetic >is based on the functions with interval parameters in the following form: > >z=f(x) e.g. [-1, 1] = sin([-10, 10]) > >or > >z=f(x,y) >e.g. >[2,5]=+([1,2],[1,3]) or [2,5]=[1,2]+[1,3] > >[1,6]=*([1,2],[1,3]) or [1,6]=[1,2]*[1,3] > >i.e. f=+,-,*,/. > >However, according to my knowledge >in most of the books which are related to interval arithmetic >there are information about systems of linear interval equations. >In this case relation between input and output interval parameters is much >more complicated. > >Y=hull{{x: Ax=B, A belong to [A], B belong to [B]} > >We have also united solution set, tolerable solution set, >controllable solution set etc. > >My question is. > >**************************************************************************** >Does theory of systems of linear interval equations belong >to interval arithmetic or not? >**************************************************************************** > >If more complicated relations on intervals >(e.g. system of linear interval equations) also belong >to interval arithmetic >then maybe all functions on intervals are some extensions >of interval arithmetic. > >I suspect that the authors of the presented book >(i.e. Uncertain Input Data Problems and the Worst Scenario Method) >work on some functions with intervals or set valued parameters >(I am not sure, I do not know that book). >Then from that point of view >they are working on some extension of interval arithmetic. > >Regards, > >Andrzej Pownuk > >--------------------------------------------- >Ph.D., research associate at: >Chair of Theoretical Mechanics >Faculty of Civil Engineering >Silesian University of Technology >ul. Krzywoustego 7 >44-100  Gliwice, Poland > >Tel/fax: 0048 32 2371542 >Mobile:  0048 606 550147 > >URL:    http://zeus.polsl.gliwice.pl/~pownuk >E-mail: pownuk [at] zeus [dot] polsl.gliwice.pl >--------------------------------------------- > >I agree. > >I have ordered it from Amazon.  Perhaps the authors don't know >about intervals.  Many people still do not. > >The encouraging thing is that even point people are starting >to realize the importance of input uncertainty.  This can only >be good for us. :) > >Cheers, > >Bill > >P.S. If nothing else, this is a source of good examples for >you to use in your talk. :) > > >Ray Moore wrote: > >> It seems odd to me that there is no mention of interval computation in >> the book. >> >> Ray Moore >> >> ----- Original Message ----- From: "Vladik Kreinovich" > >> To: >> Cc: >> Sent: Wednesday, February 16, 2005 12:27 AM >> Subject: a book of potential interest to interval researchers >> >> >>> Forwarding, from NA Digest >>> ****************************************** >>> From: Jan Chleboun >>> Date: Wed, 9 Feb 2005 12:09:50 +0100 (CET) >>> Subject: New Book on Uncertain Input Data >>> >>> Uncertain Input Data Problems and the Worst Scenario Method >>> by Ivan Hlavacek, Jan Chleboun, and Ivo Babuska; >>> North-Holland Series in Applied Mathematics and Mechanics, Volume 46, >>> Elsevier, Amsterdam, December 2004. >>> Hardback, xxvi+458 pages, ISBN: 0-444-51435-X >>> >>> This book deals with the impact of uncertainty in input data on the >>> outputs of mathematical models. Uncertain inputs as scalars, tensors, >>> functions, or domain boundaries are considered. In practical terms, >>> material parameters or constitutive laws, for instance, are uncertain, >>> and >>> quantities as local temperature, local mechanical stress, or local >>> displacement are monitored. The goal of the worst scenario method is to >>> extremize the quantity over the set of uncertain input data. >>> >>> A general mathematical scheme of the worst scenario method, including >>> approximation by finite element methods, is presented, and then >>> applied to >>> various state problems modeled by differential equations or variational >>> inequalities: nonlinear heat flow, Timoshenko beam vibration and >>> buckling, >>> plate buckling, contact problems in elasticity and thermoelasticity with >>> and without friction, and various models of plastic deformation, to list >>> some of the topics. Dozens of examples, figures, and tables are included. >>> >>> More details can be found at >>> http://books.elsevier.com/elsevier/?isbn=044451435X >>> >>> Jan Chleboun >>> Mathematical Institute >>> Academy of Sciences >>> Zitna 25  ( TeX: \v{Z}itn\'a ) >>> 115 67 Prague 1 >>> Czech Republic >>> e-mail: chleb [at] math [dot] cas.cz >>> >>> >> > > > > --------------------------------------------------------------- R. Baker Kearfott, rbk [at] louisiana [dot] edu (337) 482-5346 (fax) (337) 482-5270 (work) (337) 993-1827 (home) URL: http://interval.louisiana.edu/kearfott.html Department of Mathematics, University of Louisiana at Lafayette (Room 217 Maxim D. Doucet Hall, 1403 Johnston Street) Box 4-1010, Lafayette, LA 70504-1010, USA --------------------------------------------------------------- From owner-reliable_computing [at] interval [dot] louisiana.edu Wed Feb 16 17:03:08 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GN38OY001064 for ; Wed, 16 Feb 2005 17:03:08 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1GN38AH001063 for reliable_computing-outgoing; Wed, 16 Feb 2005 17:03:08 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1GN2xHA001059 for ; Wed, 16 Feb 2005 17:03:05 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j1GN2oN25309; Wed, 16 Feb 2005 16:02:50 -0700 (MST) Message-Id: <200502162302.j1GN2oN25309 [at] cs [dot] utep.edu> Date: Wed, 16 Feb 2005 16:02:50 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: RE: a book of potential interest to interval researchers To: reliable_computing [at] interval [dot] louisiana.edu, andrzej.pownuk [at] polsl [dot] pl MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: 29vrR+ZyduZ8XBYIVP/GBw== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear Andrzej, Just one comment to what Baker said: united solution set is EXACTLY what interval computations is about. In general, one of the main problems of interval computation is, given a function f(x1,...,xn) and intervals [x1],...,[xn], find teh range of f on these intervals. For linear equations, the inputs are coefficients of the matrix A and of the right-hand side b, and i-th function is a function that, given A and b, returns i-th component of the solution. Then the range of this function is exactly i-th component of the united solution set. This is, by the way, how we introduce linear systems in our complexity of interval computations book (written with Jiri Rohn). Vladik > From: "Andrzej Pownuk" > To: > Subject: RE: a book of potential interest to interval researchers > Date: Wed, 16 Feb 2005 22:31:37 +0100 > X-Priority: 3 (Normal) > X-MSMail-Priority: Normal > X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 > Importance: Normal > X-OriginalArrivalTime: 16 Feb 2005 21:31:33.0961 (UTC) FILETIME=[E3230B90:01C5146E] > > Dear interval researchers, > > I have general question related to the problem, > what is interval arithmetic and what is not interval arithmetic. > > As far as I know interval arithmetic > is based on the functions with interval parameters in the following form: > > z=f(x) e.g. [-1, 1] = sin([-10, 10]) > > or > > z=f(x,y) > e.g. > [2,5]=+([1,2],[1,3]) or [2,5]=[1,2]+[1,3] > > [1,6]=*([1,2],[1,3]) or [1,6]=[1,2]*[1,3] > > i.e. f=+,-,*,/. > > However, according to my knowledge > in most of the books which are related to interval arithmetic > there are information about systems of linear interval equations. > In this case relation between input and output interval parameters is much > more complicated. > > Y=hull{{x: Ax=B, A belong to [A], B belong to [B]} > > We have also united solution set, tolerable solution set, > controllable solution set etc. > > My question is. > > **************************************************************************** > Does theory of systems of linear interval equations belong > to interval arithmetic or not? > **************************************************************************** > > If more complicated relations on intervals > (e.g. system of linear interval equations) also belong > to interval arithmetic > then maybe all functions on intervals are some extensions > of interval arithmetic. > > I suspect that the authors of the presented book > (i.e. Uncertain Input Data Problems and the Worst Scenario Method) > work on some functions with intervals or set valued parameters > (I am not sure, I do not know that book). > Then from that point of view > they are working on some extension of interval arithmetic. > > Regards, > > Andrzej Pownuk From owner-reliable_computing [at] interval [dot] louisiana.edu Thu Feb 17 06:44:23 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1HCiNNi002568 for ; Thu, 17 Feb 2005 06:44:23 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1HCiNR1002567 for reliable_computing-outgoing; Thu, 17 Feb 2005 06:44:23 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1HCiKE0002563 for ; Thu, 17 Feb 2005 06:44:20 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1HCiKDE002562 for reliable_computing [at] interval [dot] louisiana.edu; Thu, 17 Feb 2005 06:44:20 -0600 (CST) Received: from scanner2.ics.uci.edu (scanner2.ics.uci.edu [128.195.1.36]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1H5F1B2001568 for ; Wed, 16 Feb 2005 23:15:10 -0600 (CST) Received: from igor.ics.uci.edu (wayne [at] igor [dot] ics.uci.edu [128.195.4.119]) by scanner2.ics.uci.edu (8.12.10/8.12.10) with ESMTP id j1H5DaAH016489 for ; Wed, 16 Feb 2005 21:13:37 -0800 (PST) Date: Wed, 16 Feb 2005 21:13:36 -0800 From: Wayne Hayes To: reliable_computing [at] interval [dot] louisiana.edu Subject: Does a more recent version of "UniCalc" exist? Message-ID: <20050217051336.GA28708 [at] igor [dot] ics.uci.edu> References: <2.2.32.20041102155117.00a19374 [at] pop [dot] louisiana.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <2.2.32.20041102155117.00a19374 [at] pop [dot] louisiana.edu> User-Agent: Mutt/1.4i X-ICS-MailScanner: Found to be clean X-ICS-MailScanner-SpamCheck: not spam (whitelisted), SpamAssassin (score=-96.69, required 5, MSGID_FROM_MTA_SHORT, USER_IN_WHITELIST) Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Hello. I've been playing with Jeff Tupper's wonderful GrafEq program (http://www.peda.com/grafeq), but I would like to be able to solve problems in more than two variables. Some searching around finds a program called UniCalc, http://archives.math.utk.edu/software/msdos/miscellaneous/unicalc/.html although it's a little archaic (from about 1995, only working under MS-DOS). Is there anything a bit more modern that combines the versatility of UniCalc with a more modern interface, and presumably updated algorithms, such as those found in GrafEq? - Wayne Hayes From owner-reliable_computing [at] interval [dot] louisiana.edu Thu Feb 17 06:45:37 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1HCjaVI002605 for ; Thu, 17 Feb 2005 06:45:36 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1HCja6h002604 for reliable_computing-outgoing; Thu, 17 Feb 2005 06:45:36 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1HCjXv2002600 for ; Thu, 17 Feb 2005 06:45:33 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1HCjXEb002599 for reliable_computing [at] interval [dot] louisiana.edu; Thu, 17 Feb 2005 06:45:33 -0600 (CST) Received: from matsrv.math.cas.cz (matsrv.math.cas.cz [147.231.88.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1H8ZG5g001847 for ; Thu, 17 Feb 2005 02:35:22 -0600 (CST) Received: from matsrv.math.cas.cz (matsrv.math.cas.cz [127.0.0.1]) by matsrv.math.cas.cz (8.12.11/8.12.11) with ESMTP id j1H8Z4pa018955; Thu, 17 Feb 2005 09:35:04 +0100 Received: from localhost (chleb@localhost) by matsrv.math.cas.cz (8.12.11/8.12.11/Submit) with ESMTP id j1H8Z019018913; Thu, 17 Feb 2005 09:35:00 +0100 Date: Thu, 17 Feb 2005 09:35:00 +0100 (CET) From: Jan Chleboun To: Vladik Kreinovich cc: Ray Moore , "G. William Walster" , reliable_computing [at] interval [dot] louisiana.edu, pardalos [at] cao [dot] ise.ufl.edu Subject: Re: a book of potential interest to interval researchers In-Reply-To: <4213796C.3080008 [at] sun [dot] com> Message-ID: References: <200502160527.j1G5RBt19348 [at] cs [dot] utep.edu> <000701c5141f$364f8550$1702a8c0@Moore> <4213796C.3080008 [at] sun [dot] com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed X-Virus-Scanned: ClamAV 0.82/707/Wed Feb 16 23:00:07 2005 on matsrv.math.cas.cz X-Virus-Status: Clean X-Spam-Status: No, hits=-4.9 required=5.0 tests=BAYES_00 autolearn=no version=2.64 X-Spam-Checker-Version: SpamAssassin 2.64 (2004-01-11) on matsrv.math.cas.cz Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear Professors Kreinovich, Moore, Pardalos, and Walster, > It seems odd to me that there is no mention of interval computation in > the book. > You are right, interval arithmetic is only briefly mentioned at page xx and page 9 of the book. However, intervals instead of crisp values appear in Timoshenko beam problems or a plasticity model, for instance. The related mathematical problems are solved analytically or numerically using standard computer arithmetic. With best regards, Jan Chleboun ****************************** Jan Chleboun Mathematical Institute Academy of Sciences Zitna 25 ( TeX: \v{Z}itn\'a ) 115 67 Prague 1 Czech Republic e-mail: chleb [at] math [dot] cas.cz ****************************** From owner-reliable_computing [at] interval [dot] louisiana.edu Thu Feb 17 08:30:13 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1HEUDeE002947 for ; Thu, 17 Feb 2005 08:30:13 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1HEUDgp002946 for reliable_computing-outgoing; Thu, 17 Feb 2005 08:30:13 -0600 (CST) Received: from zmit1.ippt.gov.pl (zmit1.ippt.gov.pl [148.81.53.8]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1HEU3A9002942 for ; Thu, 17 Feb 2005 08:30:09 -0600 (CST) Received: (from zkulpa@localhost) by zmit1.ippt.gov.pl (8.8.5/8.7.3-zmit) id PAA23799 for reliable_computing [at] interval [dot] louisiana.edu; Thu, 17 Feb 2005 15:24:39 +0100 (MET) Date: Thu, 17 Feb 2005 15:24:39 +0100 (MET) From: Zenon Kulpa Message-Id: <200502171424.PAA23799 [at] zmit1 [dot] ippt.gov.pl> To: reliable_computing [at] interval [dot] louisiana.edu Subject: RE: a book of potential interest to interval researchers Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk > From owner-reliable_computing [at] interval [dot] louisiana.edu Wed Feb 16 22:27:57 2005 > From: "Andrzej Pownuk" > [...]> > My question is. > > ****************************************************************** > Does theory of systems of linear interval equations belong > to interval arithmetic or not? > ****************************************************************** > Positively yes, I presume... [...] > I suspect that the authors of the presented book > (i.e. Uncertain Input Data Problems and the Worst Scenario Method) > work on some functions with intervals or set valued parameters > (I am not sure, I do not know that book). > Then from that point of view > they are working on some extension of interval arithmetic. > Certainly, as far as I can judge from the information posted to the list, they are certainly working on interval arithmetic. That is not the problem - the problem is, as I understand the post of prof. Moore, that they do not mention/advertise that fact (or only very accidentally, as a recent post by Jan Chleboun clarifies). -- Zenon Kulpa From owner-reliable_computing [at] interval [dot] louisiana.edu Thu Feb 17 08:38:24 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1HEcOrT003066 for ; Thu, 17 Feb 2005 08:38:24 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1HEcNjN003065 for reliable_computing-outgoing; Thu, 17 Feb 2005 08:38:23 -0600 (CST) Received: from ms-smtp-03-eri0.ohiordc.rr.com (ms-smtp-03-smtplb.ohiordc.rr.com [65.24.5.137]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1HEc9hR003061 for ; Thu, 17 Feb 2005 08:38:20 -0600 (CST) Received: from Moore (wor226111.columbus.rr.com [204.210.226.111]) by ms-smtp-03-eri0.ohiordc.rr.com (8.12.10/8.12.7) with SMTP id j1HEc2wZ009149; Thu, 17 Feb 2005 09:38:03 -0500 (EST) Message-ID: <000901c514fe$4f3fb0f0$1702a8c0@Moore> From: "Ray Moore" To: "Zenon Kulpa" , References: <200502171424.PAA23799 [at] zmit1 [dot] ippt.gov.pl> Subject: Re: a book of potential interest to interval researchers Date: Thu, 17 Feb 2005 09:38:12 -0500 MIME-Version: 1.0 Content-Type: text/plain; format=flowed; charset="iso-8859-1"; reply-type=original Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.2180 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 X-Virus-Scanned: Symantec AntiVirus Scan Engine Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Perhaps we should all read the book before we get too involved in discussion of it. We are told that we should not judge a book by its cover. Ray Moore ----- Original Message ----- From: "Zenon Kulpa" To: Sent: Thursday, February 17, 2005 9:24 AM Subject: RE: a book of potential interest to interval researchers >> From owner-reliable_computing [at] interval [dot] louisiana.edu Wed Feb 16 22:27:57 >> 2005 >> From: "Andrzej Pownuk" >> > [...]> >> My question is. >> >> ****************************************************************** >> Does theory of systems of linear interval equations belong >> to interval arithmetic or not? >> ****************************************************************** >> > Positively yes, I presume... > > [...] >> I suspect that the authors of the presented book >> (i.e. Uncertain Input Data Problems and the Worst Scenario Method) >> work on some functions with intervals or set valued parameters >> (I am not sure, I do not know that book). >> Then from that point of view >> they are working on some extension of interval arithmetic. >> > Certainly, as far as I can judge from the information posted > to the list, they are certainly working on interval arithmetic. > That is not the problem - the problem is, as I understand > the post of prof. Moore, that they do not mention/advertise > that fact (or only very accidentally, as a recent post > by Jan Chleboun clarifies). > > -- Zenon Kulpa > From owner-reliable_computing [at] interval [dot] louisiana.edu Thu Feb 17 10:15:40 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1HGFe9S003276 for ; Thu, 17 Feb 2005 10:15:40 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1HGFeAt003275 for reliable_computing-outgoing; Thu, 17 Feb 2005 10:15:40 -0600 (CST) Received: from mail-mta.sunlabs.com (dyn50.sunlabs.com [204.153.12.50]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1HGFU0u003271 for ; Thu, 17 Feb 2005 10:15:36 -0600 (CST) Received: from mail.sunlabs.com ([152.70.2.186]) by mail-mta.sfvic.sunlabs.com (Sun Java System Messaging Server 6.1 HotFix 0.02 (built Aug 25 2004)) with ESMTP id <0IC200734D5E9C00@mail-mta.sfvic.sunlabs.com> for reliable_computing [at] interval [dot] louisiana.edu; Thu, 17 Feb 2005 08:15:14 -0800 (PST) Received: from [129.146.72.246] by mail.sunlabs.com (Sun Java System Messaging Server 6.1 HotFix 0.02 (built Aug 25 2004)) with ESMTPSA id <0IC200CE2D5B8A10 [at] mail [dot] sunlabs.com> for reliable_computing [at] interval [dot] louisiana.edu; Thu, 17 Feb 2005 08:15:11 -0800 (PST) Date: Thu, 17 Feb 2005 08:15:14 -0800 From: "G. William Walster" Subject: Re: Does a more recent version of "UniCalc" exist? In-reply-to: <20050217051336.GA28708 [at] igor [dot] ics.uci.edu> To: Wayne Hayes Cc: reliable_computing [at] interval [dot] louisiana.edu Reply-to: Bill.Walster [at] sun [dot] com Message-id: <4214C312.6070304 [at] sun [dot] com> Organization: Sun Microsystems Laboratories MIME-version: 1.0 Content-type: text/plain; charset=us-ascii; format=flowed Content-transfer-encoding: 7BIT X-Accept-Language: en-us, en References: <2.2.32.20041102155117.00a19374 [at] pop [dot] louisiana.edu> <20050217051336.GA28708 [at] igor [dot] ics.uci.edu> User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7) Gecko/20040616 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Wayne, Check out Graphing Calculator at The good news is that it has a much richer interface and the algorithms use intervals. The bad news is that the algorithms are *not* rigorous. What would be cool is to have the interface and versitility of Graphing Calculator with the rigor of GrafEq. Perhaps the community could begin to lobby for such a development. Cheers, Bill Wayne Hayes wrote: > Hello. I've been playing with Jeff Tupper's wonderful GrafEq program > (http://www.peda.com/grafeq), but I would like to be able to solve > problems in more than two variables. Some searching around finds > a program called UniCalc, > > http://archives.math.utk.edu/software/msdos/miscellaneous/unicalc/.html > > although it's a little archaic (from about 1995, only working under > MS-DOS). Is there anything a bit more modern that combines the > versatility of UniCalc with a more modern interface, and presumably > updated algorithms, such as those found in GrafEq? > > - Wayne Hayes > From owner-reliable_computing [at] interval [dot] louisiana.edu Fri Feb 18 03:44:49 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1I9imfb004844 for ; Fri, 18 Feb 2005 03:44:48 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1I9imWU004843 for reliable_computing-outgoing; Fri, 18 Feb 2005 03:44:48 -0600 (CST) Received: from mx.iis.nsk.su (mxalt.iis.nsk.su [194.226.177.83]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1I9iZ5O004838 for ; Fri, 18 Feb 2005 03:44:43 -0600 (CST) Received: from [192.168.1.135] (bream.rriai.org.ru [194.226.177.135]) (authenticated bits=0) by mx.iis.nsk.su (8.12.9/8.12.9) with ESMTP id j1I9i0E8016617; Fri, 18 Feb 2005 15:44:07 +0600 (NOVT) (envelope-from pes [at] iis [dot] nsk.su) Message-ID: <4215B8E5.8090207 [at] iis [dot] nsk.su> Date: Fri, 18 Feb 2005 15:44:05 +0600 From: Evgueni Petrov User-Agent: Mozilla Thunderbird 1.0 (Windows/20041206) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Wayne Hayes CC: reliable_computing [at] interval [dot] louisiana.edu Subject: Re: Does a more recent version of "UniCalc" exist? References: <2.2.32.20041102155117.00a19374 [at] pop [dot] louisiana.edu> <20050217051336.GA28708 [at] igor [dot] ics.uci.edu> In-Reply-To: <20050217051336.GA28708 [at] igor [dot] ics.uci.edu> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Hello Wayne, Here are the facts concerning Unicalc that i am aware of. Yes, Unicalc is accessible through web-interface at http://speedy.iis.nsk.su. Please keep in mind that this site is not maintained regularly. You will find two solvers there: Unicalc3 (the one you mention) and a beta version of Unicalc5. Please see Introduction for general instructions on usage. All your runs of Unicalc will be limited to 15 variables and 5 seconds of CPU time. Hopefully, Unicalc5 will be released April-May 2005. Best regards, Evgueni Petrov. Wayne Hayes wrote: >Hello. I've been playing with Jeff Tupper's wonderful GrafEq program >(http://www.peda.com/grafeq), but I would like to be able to solve >problems in more than two variables. Some searching around finds >a program called UniCalc, > > http://archives.math.utk.edu/software/msdos/miscellaneous/unicalc/.html > >although it's a little archaic (from about 1995, only working under >MS-DOS). Is there anything a bit more modern that combines the >versatility of UniCalc with a more modern interface, and presumably >updated algorithms, such as those found in GrafEq? > > - Wayne Hayes > From owner-reliable_computing [at] interval [dot] louisiana.edu Fri Feb 18 04:40:37 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1IAean8005347 for ; Fri, 18 Feb 2005 04:40:36 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1IAeauY005346 for reliable_computing-outgoing; Fri, 18 Feb 2005 04:40:36 -0600 (CST) Received: from interferon.mpi-sb.mpg.de (mail.mpi-sb.mpg.de [139.19.1.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1IAeRZE005342 for ; Fri, 18 Feb 2005 04:40:33 -0600 (CST) Received: from amavis by interferon.mpi-sb.mpg.de with scanned-ok (Exim 3.36 #1 (Debian)) id 1D25Yb-0002tW-00 for ; Fri, 18 Feb 2005 11:40:21 +0100 Received: from mpino2303.ag2.mpi-sb.mpg.de ([139.19.24.75]) by interferon.mpi-sb.mpg.de with esmtp (Exim 3.36 #1 (Debian)) id 1D25YX-0002sW-00; Fri, 18 Feb 2005 11:40:17 +0100 Subject: Re: Does a more recent version of "UniCalc" exist? From: Stefan Ratschan To: Wayne Hayes Cc: reliable_computing [at] interval [dot] louisiana.edu In-Reply-To: <20050217051336.GA28708 [at] igor [dot] ics.uci.edu> References: <2.2.32.20041102155117.00a19374 [at] pop [dot] louisiana.edu> <20050217051336.GA28708 [at] igor [dot] ics.uci.edu> Content-Type: text/plain Date: Fri, 18 Feb 2005 11:39:46 +0100 Message-Id: <1108723187.3498.18.camel@localhost> Mime-Version: 1.0 X-Mailer: Evolution 2.0.3 Content-Transfer-Encoding: 7bit X-Virus-Scanned: by AMaViS perl-11 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk On Mit, 2005-02-16 at 21:13 -0800, Wayne Hayes wrote: > Hello. I've been playing with Jeff Tupper's wonderful GrafEq program > (http://www.peda.com/grafeq), but I would like to be able to solve > problems in more than two variables. Our solver can compute validated over- and underapproximations of the projection to 2-dimensional space of higher-dimensional solution sets of systems of inequalities: http://www.mpi-sb.mpg.de/~ratschan/rsolver/ Try to enter things like EXISTS [z] [[-2,2]] [ x^2+y^2+z^2<=1 /\ z>y ]; in the provided web interface. However, please be aware that all of this is still a research prototype. Stefan Ratschan From owner-reliable_computing [at] interval [dot] louisiana.edu Sat Feb 19 06:16:58 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1JCGvjw007465 for ; Sat, 19 Feb 2005 06:16:58 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1JCGvM5007464 for reliable_computing-outgoing; Sat, 19 Feb 2005 06:16:57 -0600 (CST) Received: from relay.pair.com (relay00.pair.com [209.68.1.20]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with SMTP id j1JCGmA2007460 for ; Sat, 19 Feb 2005 06:16:54 -0600 (CST) Date: Sat, 19 Feb 2005 06:16:49 -0600 (CST) Message-Id: <200502191216.j1JCGmA2007460 [at] interval [dot] louisiana.edu> Received: (qmail 7464 invoked from network); 19 Feb 2005 12:15:37 -0000 Received: from unknown (HELO peda.com) (unknown) by unknown with SMTP; 19 Feb 2005 12:15:37 -0000 X-pair-Authenticated: 209.68.1.35 From: Jeff Tupper To: Subject: Re: Does a more recent version of "UniCalc" exist? Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk At 9:13 PM -0800 2/16/05, Wayne Hayes wrote: >I've been playing with Jeff Tupper's wonderful GrafEq program >(http://www.peda.com/grafeq), but I would like to be able to solve >problems in more than two variables. I'm trying to write up papers on some of the graphing programs I've been working on. I've made good progress towards that end with clg (in that I've submitted a first paper on it to a conference in the U.S.), which is another two-dimensional program (which was used to produce some of the illustrations in Hansen and Walster's "Global Optimization Using Interval Analysis"; I'm giving an introductory talk that discusses some aspects of clg here in Antwerp next week [Annie Cuyt has kindly brought me over here for a few weeks] and back at Toronto in a few weeks). After I've done a bit more writing on clg, I'm hoping to bring one of my older programs that displays three-dimensional views out of hibernation. Jeff From owner-reliable_computing [at] interval [dot] louisiana.edu Mon Feb 21 08:52:42 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LEqgq0012155 for ; Mon, 21 Feb 2005 08:52:42 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LEqfLX012154 for reliable_computing-outgoing; Mon, 21 Feb 2005 08:52:41 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LEqcha012146 for ; Mon, 21 Feb 2005 08:52:38 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LEqcRL012145 for reliable_computing [at] interval [dot] louisiana.edu; Mon, 21 Feb 2005 08:52:38 -0600 (CST) Received: from scanner2.ics.uci.edu (scanner2.ics.uci.edu [128.195.1.36]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1I30hxO004301 for ; Thu, 17 Feb 2005 21:00:49 -0600 (CST) Received: from igor.ics.uci.edu (wayne [at] igor [dot] ics.uci.edu [128.195.4.119]) by scanner2.ics.uci.edu (8.12.10/8.12.10) with ESMTP id j1I2x3AH000182 for ; Thu, 17 Feb 2005 18:59:03 -0800 (PST) Date: Thu, 17 Feb 2005 18:59:03 -0800 From: Wayne Hayes To: reliable_computing [at] interval [dot] louisiana.edu Subject: Interval solver that includes complex roots (was Re: RE Does a more recent version of "UniCalc" exist?) Message-ID: <20050218025902.GA9230 [at] igor [dot] ics.uci.edu> References: <20050217051336.GA28708 [at] igor [dot] ics.uci.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.4i X-ICS-MailScanner: Found to be clean X-ICS-MailScanner-SpamCheck: not spam (whitelisted), SpamAssassin (score=-96.69, required 5, MSGID_FROM_MTA_SHORT, USER_IN_WHITELIST) Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Let me re-phrase my question. I would like to know if there exists an interval solver that includes finding complex roots. It would be sufficient if the solver solved only systems of polynomial equations, but they have to be of arbitrary degree. The application, surprisingly enough, is a branch of quantum mechanics called Yang-Mills Theory, which is a branch of what's called "string theory", related to QCD. I just saw a talk by a physicist yesterday. She has a problem that reduces to needing to know the number of roots of a system of polynomial equations. She doesn't even need to know the actual value of the roots, just how many there are. Furthermore, even a lower bound on how many roots there are would be a good start; we don't absolutely need to resolve close roots, for example, although of course it would be nice. The equations are of the form \lambda_i = C_i \sum_{k=0}^d a_k (x_i)^k, i=1,...,n 0 = \sum_{i=1}^n C_i l_i (x_i)^k , k=0,...,d and we want to solve for the x_i's. She's tried Mathematica and I think Maple, and I told her that I knew in principle that an interval solver (similar to GrafEq) should be able to solve this rigorously -- but of course it may turn out to scale exponentially in d or n. However, we cannot ignore complex roots, as they are crucial in computing quantum mechanical wave functions. I would guess that it would, in principle, not be a major step to take some existing solver and extend it to include complex roots. The question is whether it's practical. - Wayne From owner-reliable_computing [at] interval [dot] louisiana.edu Mon Feb 21 08:52:59 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LEqw9f012175 for ; Mon, 21 Feb 2005 08:52:59 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LEqwjW012174 for reliable_computing-outgoing; Mon, 21 Feb 2005 08:52:58 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LEqtii012170 for ; Mon, 21 Feb 2005 08:52:55 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LEqtdR012169 for reliable_computing [at] interval [dot] louisiana.edu; Mon, 21 Feb 2005 08:52:55 -0600 (CST) Received: from ensieta.fr (auvergne.ensieta.fr [193.52.45.6]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1HEVl3C003015 for ; Thu, 17 Feb 2005 08:31:54 -0600 (CST) Received: from ensieta.fr (localhost [127.0.0.1]) by ensieta.fr (8.9.3/jtpda-5.3.1) with ESMTP id PAA21145 for ; Thu, 17 Feb 2005 15:31:46 +0100 (MET) Received: from strauss.ensieta.ecole (strauss [172.20.1.12]) by ensieta.fr (8.9.3/jtpda-5.3.1) with ESMTP id PAA21130 ; Thu, 17 Feb 2005 15:31:44 +0100 (MET) In-Reply-To: <20050217051336.GA28708 [at] igor [dot] ics.uci.edu> To: Wayne Hayes , reliable_computing [at] interval [dot] louisiana.edu Cc: baguenar [at] istia [dot] univ-angers.fr, dao [at] istia [dot] univ-angers.fr Subject: RE Does a more recent version of "UniCalc" exist? MIME-Version: 1.0 X-Mailer: Lotus Notes Release 6.5.3 September 14, 2004 Message-ID: From: Luc JAULIN Date: Thu, 17 Feb 2005 15:29:50 +0100 X-MIMETrack: Serialize by Router on TEAM/ENSIETA(Release 6.5.3FP1 | December 22, 2004) at 17/02/2005 15:31:36, Serialize complete at 17/02/2005 15:31:36 Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by interval.louisiana.edu id j1HEVt3C003023 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear Wayne and others, You can try our windows minisolver "Proj2d" available at http://www.istia.univ-angers.fr/~dao/Proj2DV5.zip developed by Massa Dao, Xavier Baguenard and myself (as a supervisor). Proj2d draws an inner and an outer approximation of the two-dimensional projection of a set defined by nonlinear equalities or inequalities. Proj2D uses algorithms based on interval analysis and constraint propagation. Extract the zip file and run the program. You will understand quickly. You can enter equalities (such as "y=sin(x)") or inequalities (such as "sqr(x)+sin(y)+z in [1,2]"). Note that Proj2d is not yet finished and you may find some bugs. Please, if you meet any problems or if you have any suggestions, let us know. I hope that it might help you to solve your problems. Regards, Luc Luc Jaulin, Bureau D214, Laboratoire E3I2, ENSIETA 2, rue François Verny, 29806 Brest, Cedex 9 Web : http://www.ensieta.fr/e3i2/Jaulin/ Tel : +33 (0)2 98 34 89 10 Fax : +33 (0)2 98 34 87 50 Wayne Hayes Envoyé par : owner-reliable_computing [at] interval [dot] louisiana.edu 17/02/2005 06:13 A reliable_computing [at] interval [dot] louisiana.edu cc Objet Does a more recent version of "UniCalc" exist? Hello. I've been playing with Jeff Tupper's wonderful GrafEq program (http://www.peda.com/grafeq), but I would like to be able to solve problems in more than two variables. Some searching around finds a program called UniCalc, http://archives.math.utk.edu/software/msdos/miscellaneous/unicalc/.html although it's a little archaic (from about 1995, only working under MS-DOS). Is there anything a bit more modern that combines the versatility of UniCalc with a more modern interface, and presumably updated algorithms, such as those found in GrafEq? - Wayne Hayes From owner-reliable_computing [at] interval [dot] louisiana.edu Mon Feb 21 08:53:14 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LErEIf012210 for ; Mon, 21 Feb 2005 08:53:14 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LErEkv012209 for reliable_computing-outgoing; Mon, 21 Feb 2005 08:53:14 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LErAH3012205 for ; Mon, 21 Feb 2005 08:53:11 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LErAHA012204 for reliable_computing [at] interval [dot] louisiana.edu; Mon, 21 Feb 2005 08:53:10 -0600 (CST) Received: from zeus.dgp.toronto.edu (zeus.dgp.toronto.edu [128.100.4.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1IFBHAJ005767 for ; Fri, 18 Feb 2005 09:11:23 -0600 (CST) Received: from atlas.dgp.toronto.edu (atlas.dgp.toronto.edu [128.100.4.2]) by zeus.dgp.toronto.edu (Postfix) with ESMTP id D4AA07789C; Fri, 18 Feb 2005 10:11:16 -0500 (EST) Received: from atlas.dgp.toronto.edu (localhost [127.0.0.1]) by atlas.dgp.toronto.edu (8.12.8/8.12.8) with ESMTP id j1IFBGdQ007647; Fri, 18 Feb 2005 10:11:16 -0500 Received: (from mooncake@localhost) by atlas.dgp.toronto.edu (8.12.8/8.12.8/Submit) id j1IFBGkA007645; Fri, 18 Feb 2005 10:11:16 -0500 Date: Fri, 18 Feb 2005 10:11:16 -0500 From: Jeff Tupper Message-Id: <200502181511.j1IFBGkA007645 [at] atlas [dot] dgp.toronto.edu> To: reliable_computing [at] interval [dot] louisiana.edu Subject: Re: calculators Cc: tupper [at] peda [dot] com Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk >Jeff and all, > >And how could such "demand" be demonstrated ? > >Ray Moore An excellent question. I was assuming that there was no strong demand for reliability as people have become accustomed to software being unreliable (at a basic level: random crashes, poor user interface, incompatibilities, etc.). Just asking the questions displays a more pro-active stance; answers could guide future proselytization. Jeff  From owner-reliable_computing [at] interval [dot] louisiana.edu Mon Feb 21 08:53:37 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LEraa2012253 for ; Mon, 21 Feb 2005 08:53:37 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LEraXP012252 for reliable_computing-outgoing; Mon, 21 Feb 2005 08:53:36 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LErQKr012234 for ; Mon, 21 Feb 2005 08:53:26 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LErQ71012233 for reliable_computing [at] interval [dot] louisiana.edu; Mon, 21 Feb 2005 08:53:26 -0600 (CST) Received: from zeus.dgp.toronto.edu (zeus.dgp.toronto.edu [128.100.4.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1IFV0qV005822 for ; Fri, 18 Feb 2005 09:31:06 -0600 (CST) Received: from atlas.dgp.toronto.edu (atlas.dgp.toronto.edu [128.100.4.2]) by zeus.dgp.toronto.edu (Postfix) with ESMTP id B942B7789F; Fri, 18 Feb 2005 10:31:00 -0500 (EST) Received: from atlas.dgp.toronto.edu (localhost [127.0.0.1]) by atlas.dgp.toronto.edu (8.12.8/8.12.8) with ESMTP id j1IFV0dQ007703; Fri, 18 Feb 2005 10:31:00 -0500 Received: (from mooncake@localhost) by atlas.dgp.toronto.edu (8.12.8/8.12.8/Submit) id j1IFV0bW007701; Fri, 18 Feb 2005 10:31:00 -0500 Date: Fri, 18 Feb 2005 10:31:00 -0500 From: Jeff Tupper Message-Id: <200502181531.j1IFV0bW007701 [at] atlas [dot] dgp.toronto.edu> To: reliable_computing [at] interval [dot] louisiana.edu Subject: Re: Does a more recent version of "UniCalc" exist? Cc: tupper [at] peda [dot] com Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk At 9:13 PM -0800 2/16/05, Wayne Hayes wrote: >I've been playing with Jeff Tupper's wonderful GrafEq program >(http://www.peda.com/grafeq), but I would like to be able to solve >problems in more than two variables. I'm trying to write up papers on some of the graphing programs I've been working on. I've made good progress towards that end with clg, which is another two-dimensional program (which was used to produce some of the illustrations in Hansen and Walster's "Global Optimization Using Interval Analysis"; I'm giving an introductory talk that discusses some aspects of clg here in Antwerp next week [Annie Cuyt has kindly brought me over here for a few weeks] and back at Toronto in a few weeks). After I've done a bit more writing on clg, I'm hoping to bring one of my older programs that displays three-dimensional views out of hibernation. Jeff From owner-reliable_computing [at] interval [dot] louisiana.edu Mon Feb 21 08:53:56 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LErt1u012297 for ; Mon, 21 Feb 2005 08:53:56 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LErsv6012293 for reliable_computing-outgoing; Mon, 21 Feb 2005 08:53:54 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LErfvR012265 for ; Mon, 21 Feb 2005 08:53:41 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LErehA012262 for reliable_computing [at] interval [dot] louisiana.edu; Mon, 21 Feb 2005 08:53:40 -0600 (CST) Received: from ns.ict.nsk.su (ns.ict.nsk.su [193.124.243.33]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1I89C7U004708 for ; Fri, 18 Feb 2005 02:09:21 -0600 (CST) Received: from COMP0 (COMP0.ict.local.net [192.168.0.124]) by ns.ict.nsk.su (8.13.1/8.13.1) with SMTP id j1I88vSj001758; Fri, 18 Feb 2005 14:08:57 +0600 (NOVT) (envelope-from shary [at] ict [dot] nsc.ru) X-Authentication-Warning: ns.ict.nsk.su: Host COMP0.ict.local.net [192.168.0.124] claimed to be COMP0 Message-ID: <003001c51591$4c979190$7c00a8c0@COMP0> From: "Sergey P. Shary" To: Cc: =?koi8-r?B?4i7zLiDk1sHO2cLFy8/X?= Subject: INTLAB problems Date: Fri, 18 Feb 2005 14:10:24 +0600 Organization: ICT SB RAS MIME-Version: 1.0 Content-Type: text/plain; charset="koi8-r" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1106 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Colleagues, I have a question on how INTLAB and numerous INTLAB codes operate interval matrices. Running a simple INTLAB function ========================================== function y = a_bug a = [ infsup(0,1) infsup(2,3) ]; b = [ infsup(-2,0.5) infsup(1,3) ]; y = a*b'; %y = 0; %for k = 1 : 2; % y = y + a(k)*b(k); %end =========================================== that computes a "scalar product" of two interval 2-vectors, I get the answer >> infsup(a_bug) intval = [ -1.00000000000000, 10.25000000000000] However, rewriting the above in mathematically equivalent algorithm that uses "for" cycle ============================================== function y = a_bug a = [ infsup(0,1) infsup(2,3) ]; b = [ infsup(-2,0.5) infsup(1,3) ]; %y = a*b'; y = 0; for k = 1 : 2; y = y + a(k)*b(k); end ============================================== produces >> infsup(a_bug) intval = [ 0.00000000000000, 9.50000000000000] What is the reason of such a discrepancy? It seems to be well known: INTLAB uses a special "fast" midpoint-radius algorithm for interval matrix multiplication called in compact MATLAB form, and this algorithm is not "sharp". Well, then I do not understand why G. Hargreaves in his M.S. thesis "Interval analysis in MATLAB" as well as many others use such a compact form of the interval matrix product in their INTLAB codes, thus getting a substantial overestimation of their results? This is the case, in particular, for interval Gauss method and Krawczyk method whose codes are written out by Hargreaves in his thesis. Sergey P. Shary From owner-reliable_computing [at] interval [dot] louisiana.edu Mon Feb 21 09:22:49 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LFMnbo012602 for ; Mon, 21 Feb 2005 09:22:49 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LFMmJq012601 for reliable_computing-outgoing; Mon, 21 Feb 2005 09:22:48 -0600 (CST) Received: from smtp2.rz.tu-harburg.de (smtp2.rz.tu-harburg.de [134.28.205.13]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LFMdFj012597 for ; Mon, 21 Feb 2005 09:22:45 -0600 (CST) Received: from mail2.rz.tu-harburg.de (mail2.rz.tu-harburg.de [134.28.202.179]) by smtp2.rz.tu-harburg.de (8.13.1/8.13.1) with ESMTP id j1LFLnfd014674 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK); Mon, 21 Feb 2005 16:21:50 +0100 Received: from oem-9n90y5vdt8b.net.waseda.ac.jp ([133.9.117.168]) (user=ti3sr mech=LOGIN bits=0) by mail2.rz.tu-harburg.de (8.13.1/8.13.1) with ESMTP id j1LFLgFx023085 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NO); Mon, 21 Feb 2005 16:21:47 +0100 To: "Sergey P. Shary" , reliable_computing [at] interval [dot] louisiana.edu Cc: =?utf-8?B?0JEu0KEuINCU0LbQsNC90YvQsdC10LrQvtCy?= Subject: Re: INTLAB problems References: <003001c51591$4c979190$7c00a8c0@COMP0> Message-ID: From: "Siegfried M. Rump" Organization: Technical University Hamburg-Harburg Content-Type: text/plain; format=flowed; delsp=yes; charset=iso-8859-15 MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Date: Mon, 21 Feb 2005 16:21:40 -0100 In-Reply-To: <003001c51591$4c979190$7c00a8c0@COMP0> User-Agent: Opera M2/7.54 (Win32, build 3869) X-Scanned-By: TUHH Rechenzentrum content checker on 134.28.205.13 X-Scanned-By: TUHH Rechenzentrum content checker on 134.28.202.179 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Sergey, you can change that: >> intvalinit('FastIVmult'), a=[infsup(0,1) infsup(2,3)]; >> b=[infsup(-2,0.5) infsup(1,3)]; a*b' ===> Fast interval matrix multiplication in use (maximum overestimation factor 1.5 in radius) intval ans = [ -1.00000000000000, 10.25000000000001] >> intvalinit('SharpIVmult'), a=[infsup(0,1) infsup(2,3)]; >> b=[infsup(-2,0.5) infsup(1,3)]; a*b' ===> Slow but sharp interval matrix multiplication in use intval ans = [ 0.00000000000000, 9.50000000000000] >> Best wishes Siegfried On Fri, 18 Feb 2005 14:10:24 +0600, Sergey P. Shary wrote: > > > > > Colleagues, > > > > > > I have a question on how INTLAB and numerous INTLAB codes operate > interval matrices. > > > > > Running a simple INTLAB function > > ========================================== > > function y = a_bug > > a = [ infsup(0,1) infsup(2,3) ]; > > b = [ infsup(-2,0.5) infsup(1,3) ]; > > > y = a*b'; > > > %y = 0; > > %for k = 1 : 2; > > % y = y + a(k)*b(k); > > %end > > > =========================================== > > > that computes a "scalar product" of two interval 2-vectors, > > I get the answer > > > >> infsup(a_bug) > > intval = > > [ -1.00000000000000, 10.25000000000000] > > > > > However, rewriting the above in mathematically equivalent algorithm > > that uses "for" cycle > > > ============================================== > > > function y = a_bug > > > a = [ infsup(0,1) infsup(2,3) ]; > > b = [ infsup(-2,0.5) infsup(1,3) ]; > > > %y = a*b'; > > > y = 0; > > for k = 1 : 2; > > y = y + a(k)*b(k); > > end > > > ============================================== > > > > produces > > > >> infsup(a_bug) > > intval = > > [ 0.00000000000000, 9.50000000000000] > > > > > > > What is the reason of such a discrepancy? > > It seems to be well known: INTLAB uses a special "fast" midpoint-radius > algorithm > > for interval matrix multiplication called in compact MATLAB form, and > this algorithm > > is not "sharp". > > > > Well, then I do not understand why G. Hargreaves in his M.S. thesis > "Interval analysis in MATLAB" > > as well as many others use such a compact form of the interval matrix > product in their INTLAB codes, > > thus getting a substantial overestimation of their results? > > > > This is the case, in particular, for interval Gauss method and Krawczyk > method > > whose codes are written out by Hargreaves in his thesis. > > > > > > Sergey P. Shary > > > > > > > -- ================================================= Prof. Dr. Siegfried M. Rump Arbeitsbereich Informatik III Technische Universität Hamburg-Harburg Schwarzenbergstr. 95 21071 Hamburg Germany phone +49 40 42878 3027 secr. +49 40 42878 3227 fax +49 40 42878 2489 From owner-reliable_computing [at] interval [dot] louisiana.edu Mon Feb 21 11:59:39 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LHxdjC012980 for ; Mon, 21 Feb 2005 11:59:39 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LHxclY012979 for reliable_computing-outgoing; Mon, 21 Feb 2005 11:59:38 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LHxZ84012975 for ; Mon, 21 Feb 2005 11:59:35 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LHxYGi012974 for reliable_computing [at] interval [dot] louisiana.edu; Mon, 21 Feb 2005 11:59:34 -0600 (CST) Received: from zeus.dgp.toronto.edu (zeus.dgp.toronto.edu [128.100.4.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1IFV0qV005822 for ; Fri, 18 Feb 2005 09:31:06 -0600 (CST) Received: from atlas.dgp.toronto.edu (atlas.dgp.toronto.edu [128.100.4.2]) by zeus.dgp.toronto.edu (Postfix) with ESMTP id B942B7789F; Fri, 18 Feb 2005 10:31:00 -0500 (EST) Received: from atlas.dgp.toronto.edu (localhost [127.0.0.1]) by atlas.dgp.toronto.edu (8.12.8/8.12.8) with ESMTP id j1IFV0dQ007703; Fri, 18 Feb 2005 10:31:00 -0500 Received: (from mooncake@localhost) by atlas.dgp.toronto.edu (8.12.8/8.12.8/Submit) id j1IFV0bW007701; Fri, 18 Feb 2005 10:31:00 -0500 Date: Fri, 18 Feb 2005 10:31:00 -0500 From: Jeff Tupper Message-Id: <200502181531.j1IFV0bW007701 [at] atlas [dot] dgp.toronto.edu> To: reliable_computing [at] interval [dot] louisiana.edu Subject: Re: Does a more recent version of "UniCalc" exist? Cc: tupper [at] peda [dot] com Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk At 9:13 PM -0800 2/16/05, Wayne Hayes wrote: >I've been playing with Jeff Tupper's wonderful GrafEq program >(http://www.peda.com/grafeq), but I would like to be able to solve >problems in more than two variables. I'm trying to write up papers on some of the graphing programs I've been working on. I've made good progress towards that end with clg, which is another two-dimensional program (which was used to produce some of the illustrations in Hansen and Walster's "Global Optimization Using Interval Analysis"; I'm giving an introductory talk that discusses some aspects of clg here in Antwerp next week [Annie Cuyt has kindly brought me over here for a few weeks] and back at Toronto in a few weeks). After I've done a bit more writing on clg, I'm hoping to bring one of my older programs that displays three-dimensional views out of hibernation. Jeff From owner-reliable_computing [at] interval [dot] louisiana.edu Mon Feb 21 11:59:33 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LHxXpr012971 for ; Mon, 21 Feb 2005 11:59:33 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LHxXv5012970 for reliable_computing-outgoing; Mon, 21 Feb 2005 11:59:33 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LHxTrP012966 for ; Mon, 21 Feb 2005 11:59:29 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LHxTPW012965 for reliable_computing [at] interval [dot] louisiana.edu; Mon, 21 Feb 2005 11:59:29 -0600 (CST) Received: from zeus.dgp.toronto.edu (zeus.dgp.toronto.edu [128.100.4.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1IFBHAJ005767 for ; Fri, 18 Feb 2005 09:11:23 -0600 (CST) Received: from atlas.dgp.toronto.edu (atlas.dgp.toronto.edu [128.100.4.2]) by zeus.dgp.toronto.edu (Postfix) with ESMTP id D4AA07789C; Fri, 18 Feb 2005 10:11:16 -0500 (EST) Received: from atlas.dgp.toronto.edu (localhost [127.0.0.1]) by atlas.dgp.toronto.edu (8.12.8/8.12.8) with ESMTP id j1IFBGdQ007647; Fri, 18 Feb 2005 10:11:16 -0500 Received: (from mooncake@localhost) by atlas.dgp.toronto.edu (8.12.8/8.12.8/Submit) id j1IFBGkA007645; Fri, 18 Feb 2005 10:11:16 -0500 Date: Fri, 18 Feb 2005 10:11:16 -0500 From: Jeff Tupper Message-Id: <200502181511.j1IFBGkA007645 [at] atlas [dot] dgp.toronto.edu> To: reliable_computing [at] interval [dot] louisiana.edu Subject: Re: calculators Cc: tupper [at] peda [dot] com Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk >Jeff and all, > >And how could such "demand" be demonstrated ? > >Ray Moore An excellent question. I was assuming that there was no strong demand for reliability as people have become accustomed to software being unreliable (at a basic level: random crashes, poor user interface, incompatibilities, etc.). Just asking the questions displays a more pro-active stance; answers could guide future proselytization. Jeff  From owner-reliable_computing [at] interval [dot] louisiana.edu Mon Feb 21 11:59:50 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LHxoo7012989 for ; Mon, 21 Feb 2005 11:59:50 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LHxn2A012988 for reliable_computing-outgoing; Mon, 21 Feb 2005 11:59:49 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LHxidf012984 for ; Mon, 21 Feb 2005 11:59:44 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LHxhuL012983 for reliable_computing [at] interval [dot] louisiana.edu; Mon, 21 Feb 2005 11:59:43 -0600 (CST) Received: from ns.ict.nsk.su (ns.ict.nsk.su [193.124.243.33]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1I89C7U004708 for ; Fri, 18 Feb 2005 02:09:21 -0600 (CST) Received: from COMP0 (COMP0.ict.local.net [192.168.0.124]) by ns.ict.nsk.su (8.13.1/8.13.1) with SMTP id j1I88vSj001758; Fri, 18 Feb 2005 14:08:57 +0600 (NOVT) (envelope-from shary [at] ict [dot] nsc.ru) X-Authentication-Warning: ns.ict.nsk.su: Host COMP0.ict.local.net [192.168.0.124] claimed to be COMP0 Message-ID: <003001c51591$4c979190$7c00a8c0@COMP0> From: "Sergey P. Shary" To: Cc: =?koi8-r?B?4i7zLiDk1sHO2cLFy8/X?= Subject: INTLAB problems Date: Fri, 18 Feb 2005 14:10:24 +0600 Organization: ICT SB RAS MIME-Version: 1.0 Content-Type: text/plain; charset="koi8-r" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1106 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Colleagues, I have a question on how INTLAB and numerous INTLAB codes operate interval matrices. Running a simple INTLAB function ========================================== function y = a_bug a = [ infsup(0,1) infsup(2,3) ]; b = [ infsup(-2,0.5) infsup(1,3) ]; y = a*b'; %y = 0; %for k = 1 : 2; % y = y + a(k)*b(k); %end =========================================== that computes a "scalar product" of two interval 2-vectors, I get the answer >> infsup(a_bug) intval = [ -1.00000000000000, 10.25000000000000] However, rewriting the above in mathematically equivalent algorithm that uses "for" cycle ============================================== function y = a_bug a = [ infsup(0,1) infsup(2,3) ]; b = [ infsup(-2,0.5) infsup(1,3) ]; %y = a*b'; y = 0; for k = 1 : 2; y = y + a(k)*b(k); end ============================================== produces >> infsup(a_bug) intval = [ 0.00000000000000, 9.50000000000000] What is the reason of such a discrepancy? It seems to be well known: INTLAB uses a special "fast" midpoint-radius algorithm for interval matrix multiplication called in compact MATLAB form, and this algorithm is not "sharp". Well, then I do not understand why G. Hargreaves in his M.S. thesis "Interval analysis in MATLAB" as well as many others use such a compact form of the interval matrix product in their INTLAB codes, thus getting a substantial overestimation of their results? This is the case, in particular, for interval Gauss method and Krawczyk method whose codes are written out by Hargreaves in his thesis. Sergey P. Shary From owner-reliable_computing [at] interval [dot] louisiana.edu Mon Feb 21 11:59:21 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LHxKe4012953 for ; Mon, 21 Feb 2005 11:59:21 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LHxKK1012952 for reliable_computing-outgoing; Mon, 21 Feb 2005 11:59:20 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LHxIiY012948 for ; Mon, 21 Feb 2005 11:59:18 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LHxHgL012947 for reliable_computing [at] interval [dot] louisiana.edu; Mon, 21 Feb 2005 11:59:17 -0600 (CST) Received: from scanner2.ics.uci.edu (scanner2.ics.uci.edu [128.195.1.36]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1I30hxO004301 for ; Thu, 17 Feb 2005 21:00:49 -0600 (CST) Received: from igor.ics.uci.edu (wayne [at] igor [dot] ics.uci.edu [128.195.4.119]) by scanner2.ics.uci.edu (8.12.10/8.12.10) with ESMTP id j1I2x3AH000182 for ; Thu, 17 Feb 2005 18:59:03 -0800 (PST) Date: Thu, 17 Feb 2005 18:59:03 -0800 From: Wayne Hayes To: reliable_computing [at] interval [dot] louisiana.edu Subject: Interval solver that includes complex roots (was Re: RE Does a more recent version of "UniCalc" exist?) Message-ID: <20050218025902.GA9230 [at] igor [dot] ics.uci.edu> References: <20050217051336.GA28708 [at] igor [dot] ics.uci.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.4i X-ICS-MailScanner: Found to be clean X-ICS-MailScanner-SpamCheck: not spam (whitelisted), SpamAssassin (score=-96.69, required 5, MSGID_FROM_MTA_SHORT, USER_IN_WHITELIST) Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Let me re-phrase my question. I would like to know if there exists an interval solver that includes finding complex roots. It would be sufficient if the solver solved only systems of polynomial equations, but they have to be of arbitrary degree. The application, surprisingly enough, is a branch of quantum mechanics called Yang-Mills Theory, which is a branch of what's called "string theory", related to QCD. I just saw a talk by a physicist yesterday. She has a problem that reduces to needing to know the number of roots of a system of polynomial equations. She doesn't even need to know the actual value of the roots, just how many there are. Furthermore, even a lower bound on how many roots there are would be a good start; we don't absolutely need to resolve close roots, for example, although of course it would be nice. The equations are of the form \lambda_i = C_i \sum_{k=0}^d a_k (x_i)^k, i=1,...,n 0 = \sum_{i=1}^n C_i l_i (x_i)^k , k=0,...,d and we want to solve for the x_i's. She's tried Mathematica and I think Maple, and I told her that I knew in principle that an interval solver (similar to GrafEq) should be able to solve this rigorously -- but of course it may turn out to scale exponentially in d or n. However, we cannot ignore complex roots, as they are crucial in computing quantum mechanical wave functions. I would guess that it would, in principle, not be a major step to take some existing solver and extend it to include complex roots. The question is whether it's practical. - Wayne From owner-reliable_computing [at] interval [dot] louisiana.edu Mon Feb 21 12:42:18 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LIgIUw013084 for ; Mon, 21 Feb 2005 12:42:18 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LIgHe1013083 for reliable_computing-outgoing; Mon, 21 Feb 2005 12:42:17 -0600 (CST) Received: from marnier.ucs.louisiana.edu (root [at] marnier [dot] ucs.louisiana.edu [130.70.132.233]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LIg9Od013079 for ; Mon, 21 Feb 2005 12:42:14 -0600 (CST) Received: from Liberty (h158065.louisiana.edu [130.70.158.65]) by marnier.ucs.louisiana.edu (8.13.1/8.13.1/ull-ucs-mx-host_1.9) with SMTP id j1LIg3Ba015471 for ; Mon, 21 Feb 2005 12:42:09 -0600 (CST) Message-Id: <2.2.32.20050221184204.00e68168 [at] pop [dot] louisiana.edu> X-Sender: rbk5287 [at] pop [dot] louisiana.edu X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Mon, 21 Feb 2005 12:42:04 -0600 To: reliable_computing [at] interval [dot] louisiana.edu From: "R. Baker Kearfott" Subject: reliable_computing: Message from the list maintainer Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear colleagues, We have been having some problems with the "reliable_computing" list server. In particular, sometimes but not always, items posted to the list were received by some, but not all, members of the list. In response, we have just replaced our "bulk mailer" with "sendmail" in the process of sending the messages out. I have "approved" five messages (from addresses the list server did not recognize as being members) twice: once about 6 hours ago and once less than an hour ago (from the time on this message). I did not receive them the first time, but I received them the most recent time. Hopefully, the problem is now fixed, but time will tell. Best regards, Baker --------------------------------------------------------------- R. Baker Kearfott, rbk [at] louisiana [dot] edu (337) 482-5346 (fax) (337) 482-5270 (work) (337) 993-1827 (home) URL: http://interval.louisiana.edu/kearfott.html Department of Mathematics, University of Louisiana at Lafayette (Room 217 Maxim D. Doucet Hall, 1403 Johnston Street) Box 4-1010, Lafayette, LA 70504-1010, USA --------------------------------------------------------------- From owner-reliable_computing [at] interval [dot] louisiana.edu Mon Feb 21 11:59:27 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LHxRUE012962 for ; Mon, 21 Feb 2005 11:59:27 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LHxRVT012961 for reliable_computing-outgoing; Mon, 21 Feb 2005 11:59:27 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LHxOoG012957 for ; Mon, 21 Feb 2005 11:59:24 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1LHxNDi012956 for reliable_computing [at] interval [dot] louisiana.edu; Mon, 21 Feb 2005 11:59:23 -0600 (CST) Received: from ensieta.fr (auvergne.ensieta.fr [193.52.45.6]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1HEVl3C003015 for ; Thu, 17 Feb 2005 08:31:54 -0600 (CST) Received: from ensieta.fr (localhost [127.0.0.1]) by ensieta.fr (8.9.3/jtpda-5.3.1) with ESMTP id PAA21145 for ; Thu, 17 Feb 2005 15:31:46 +0100 (MET) Received: from strauss.ensieta.ecole (strauss [172.20.1.12]) by ensieta.fr (8.9.3/jtpda-5.3.1) with ESMTP id PAA21130 ; Thu, 17 Feb 2005 15:31:44 +0100 (MET) In-Reply-To: <20050217051336.GA28708 [at] igor [dot] ics.uci.edu> To: Wayne Hayes , reliable_computing [at] interval [dot] louisiana.edu Cc: baguenar [at] istia [dot] univ-angers.fr, dao [at] istia [dot] univ-angers.fr Subject: RE Does a more recent version of "UniCalc" exist? MIME-Version: 1.0 X-Mailer: Lotus Notes Release 6.5.3 September 14, 2004 Message-ID: From: Luc JAULIN Date: Thu, 17 Feb 2005 15:29:50 +0100 X-MIMETrack: Serialize by Router on TEAM/ENSIETA(Release 6.5.3FP1 | December 22, 2004) at 17/02/2005 15:31:36, Serialize complete at 17/02/2005 15:31:36 Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by interval.louisiana.edu id j1HEVt3C003023 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear Wayne and others, You can try our windows minisolver "Proj2d" available at http://www.istia.univ-angers.fr/~dao/Proj2DV5.zip developed by Massa Dao, Xavier Baguenard and myself (as a supervisor). Proj2d draws an inner and an outer approximation of the two-dimensional projection of a set defined by nonlinear equalities or inequalities. Proj2D uses algorithms based on interval analysis and constraint propagation. Extract the zip file and run the program. You will understand quickly. You can enter equalities (such as "y=sin(x)") or inequalities (such as "sqr(x)+sin(y)+z in [1,2]"). Note that Proj2d is not yet finished and you may find some bugs. Please, if you meet any problems or if you have any suggestions, let us know. I hope that it might help you to solve your problems. Regards, Luc Luc Jaulin, Bureau D214, Laboratoire E3I2, ENSIETA 2, rue François Verny, 29806 Brest, Cedex 9 Web : http://www.ensieta.fr/e3i2/Jaulin/ Tel : +33 (0)2 98 34 89 10 Fax : +33 (0)2 98 34 87 50 Wayne Hayes Envoyé par : owner-reliable_computing [at] interval [dot] louisiana.edu 17/02/2005 06:13 A reliable_computing [at] interval [dot] louisiana.edu cc Objet Does a more recent version of "UniCalc" exist? Hello. I've been playing with Jeff Tupper's wonderful GrafEq program (http://www.peda.com/grafeq), but I would like to be able to solve problems in more than two variables. Some searching around finds a program called UniCalc, http://archives.math.utk.edu/software/msdos/miscellaneous/unicalc/.html although it's a little archaic (from about 1995, only working under MS-DOS). Is there anything a bit more modern that combines the versatility of UniCalc with a more modern interface, and presumably updated algorithms, such as those found in GrafEq? - Wayne Hayes From owner-reliable_computing [at] interval [dot] louisiana.edu Mon Feb 21 19:54:40 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1M1se5E013561 for ; Mon, 21 Feb 2005 19:54:40 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1M1seOu013560 for reliable_computing-outgoing; Mon, 21 Feb 2005 19:54:40 -0600 (CST) Received: from mail.your.biglobe.net (mdsv0021.ht.necidc.net [202.225.207.136]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with SMTP id j1M1sUxp013556 for ; Mon, 21 Feb 2005 19:54:36 -0600 (CST) Received: (biglobe-qmail 11291 invoked by uid 0); 22 Feb 2005 10:54:16 +0900 Received: from kuma.waseda.jp [172.23.46.20] by biglobe-qmail with SMTP; 22 Feb 2005 10:54:16 +0900 Message-Id: <200502220154.AA00576 [at] kuma [dot] waseda.jp> From: Takeshi OGITA Date: Tue, 22 Feb 2005 10:54:13 +0900 To: "Sergey P. Shary" Cc: Subject: Re: INTLAB problems In-Reply-To: <003001c51591$4c979190$7c00a8c0@COMP0> References: <003001c51591$4c979190$7c00a8c0@COMP0> MIME-Version: 1.0 X-Mailer: AL-Mail32 Version 1.13 Content-Type: text/plain; charset=us-ascii Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk You can see the option by >> help intvalinit and then find Default thick real interval times thick real interval: Best regards, Takeshi Ogita > > > >Colleagues, > > > > > >I have a question on how INTLAB and numerous INTLAB codes operate interval matrices. > > > > >Running a simple INTLAB function > > ========================================== > >function y = a_bug > >a = [ infsup(0,1) infsup(2,3) ]; > >b = [ infsup(-2,0.5) infsup(1,3) ]; > > > y = a*b'; > > > %y = 0; > > %for k = 1 : 2; > > % y = y + a(k)*b(k); > > %end > > >=========================================== > > >that computes a "scalar product" of two interval 2-vectors, > >I get the answer > > > >> infsup(a_bug) > > intval = > > [ -1.00000000000000, 10.25000000000000] > > > > >However, rewriting the above in mathematically equivalent algorithm > >that uses "for" cycle > > >============================================== > > >function y = a_bug > > >a = [ infsup(0,1) infsup(2,3) ]; > >b = [ infsup(-2,0.5) infsup(1,3) ]; > > > %y = a*b'; > > > y = 0; > > for k = 1 : 2; > > y = y + a(k)*b(k); > > end > > >============================================== > > > >produces > > > >> infsup(a_bug) > > intval = > > [ 0.00000000000000, 9.50000000000000] > > > > > > >What is the reason of such a discrepancy? > >It seems to be well known: INTLAB uses a special "fast" midpoint-radius algorithm > >for interval matrix multiplication called in compact MATLAB form, and this algorithm > >is not "sharp". > > > >Well, then I do not understand why G. Hargreaves in his M.S. thesis "Interval analysis in MATLAB" > >as well as many others use such a compact form of the interval matrix product in their INTLAB codes, > >thus getting a substantial overestimation of their results? > > > >This is the case, in particular, for interval Gauss method and Krawczyk method > >whose codes are written out by Hargreaves in his thesis. > > > > > >Sergey P. Shary > > > > > > > > From owner-reliable_computing [at] interval [dot] louisiana.edu Tue Feb 22 03:03:43 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1M93g4W014096 for ; Tue, 22 Feb 2005 03:03:42 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1M93g6F014095 for reliable_computing-outgoing; Tue, 22 Feb 2005 03:03:42 -0600 (CST) Received: from sophia.inria.fr (sophia.inria.fr [138.96.64.20]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1M93Wvw014091 for ; Tue, 22 Feb 2005 03:03:38 -0600 (CST) Received: from localhost (localhost [127.0.0.1]) by sophia.inria.fr (8.12.10/8.12.9) with ESMTP id j1M93Q18025819 for ; Tue, 22 Feb 2005 10:03:26 +0100 Received: from caphorn.inria.fr (caphorn.inria.fr [138.96.116.27]) by sophia.inria.fr (8.12.10/8.12.9) with ESMTP id j1M930hh025283 for ; Tue, 22 Feb 2005 10:03:02 +0100 Received: (from merlet@localhost) by caphorn.inria.fr (8.12.10/8.12.5) id j1M92x6q012367; Tue, 22 Feb 2005 10:02:59 +0100 (MET) Date: Tue, 22 Feb 2005 10:02:59 +0100 (MET) From: Jean-Pierre Merlet Message-Id: <200502220902.j1M92x6q012367 [at] caphorn [dot] inria.fr> To: reliable_computing [at] interval [dot] louisiana.edu Subject: Re: Interval solver that includes complex roots (was Re: RE Does a more recent version of "UniCalc" exist?) In-Reply-To: Mail from 'Wayne Hayes ' dated: Thu, 17 Feb 2005 18:59:03 -0800 Cc: Jean-Pierre.Merlet [at] sophia [dot] inria.fr X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-1.4 (sophia.inria.fr [138.96.64.20]); Tue, 22 Feb 2005 10:03:02 +0100 (MET) X-Virus-Scanned: by amavisd-new at sophia.inria.fr Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Wayne, We have a Maple interface to our interval analysis library ALIAS that may allow to deal with your problem (for your information we have already worked in the field of quantum mechanics, see Mladen Pavicic et al 2005 J. Phys. A: Math. Gen. 38 1577-159, issue of February 2005: this was an interesting work for us where we have to solve systems of between 40 and 200 equations and a huge number of them, last total was about 400 millions!). Alternatively we have close contact with people working in algebraic geometry with Groebner basis and they may have a solution too. So the best way is to send us a typical example and we will see what happen. Best Jean-Pierre From owner-reliable_computing [at] interval [dot] louisiana.edu Tue Feb 22 08:22:34 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1MEMXt5014896 for ; Tue, 22 Feb 2005 08:22:34 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1MEMXYw014894 for reliable_computing-outgoing; Tue, 22 Feb 2005 08:22:33 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1MEMQLP014876 for ; Tue, 22 Feb 2005 08:22:26 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1MEMP6E014873 for reliable_computing [at] interval [dot] louisiana.edu; Tue, 22 Feb 2005 08:22:25 -0600 (CST) Received: from scanner2.ics.uci.edu (root [at] scanner2 [dot] ics.uci.edu [128.195.1.36]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LKVXhH013217 for ; Mon, 21 Feb 2005 14:31:39 -0600 (CST) Received: from igor.ics.uci.edu (igor.ics.uci.edu [128.195.4.119]) by scanner2.ics.uci.edu (8.12.10/8.12.10) with ESMTP id j1LKUpAH001701 for ; Mon, 21 Feb 2005 12:30:51 -0800 (PST) Date: Mon, 21 Feb 2005 12:30:49 -0800 From: Wayne Hayes To: reliable_computing [at] interval [dot] louisiana.edu Subject: Re: Interval solver that includes complex roots Message-ID: <20050221203049.GB16904 [at] igor [dot] ics.uci.edu> References: <20050217051336.GA28708 [at] igor [dot] ics.uci.edu> <20050218025902.GA9230 [at] igor [dot] ics.uci.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.4i X-ICS-MailScanner: Found to be clean X-ICS-MailScanner-SpamCheck: not spam (whitelisted), SpamAssassin (score=-96.69, required 5, MSGID_FROM_MTA_SHORT, USER_IN_WHITELIST) Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk On Mon, Feb 21, 2005 at 02:55:02PM -0500, Youdong Lin wrote: > We could let x_j = a_j + ib_j, where a_j and b_j are both real. > In this way, we still can use solvers for systems of > equations to find all REAL roots of the resulting 2n by 2n system, > instead of the original n by n system. Yes, this is true and is easy to do, assuming we have a solver than can solve a 2n by 2n system. Does such a solver currently exist? I know it can be done and I know how to do it, but it would be alot of work to actually code it up. Certainly this must have been done already? GrafEq would be perfect except it only works in two variables, whereas I need to solve systems of polynomial equations with an arbitrary number of equations and variables. UniCalc appears to have the language to define the problem nicely, but is not rigorous and returns answers which are only approximate, and will often claim to find multiple nearby roots where in fact only one actually exists. INTLAB looks like it might be up to the task, but I am not very famaliar to MATLAB and am reluctant to expend the effort to learn it unless I know there's a decent chance I can get it to work. GlobSol also looks promising. There are plenty of packages listed at http://www.cs.utep.edu/interval-comp/intsoft.html but I just want some hints as to were to start! :-) - Wayne From owner-reliable_computing [at] interval [dot] louisiana.edu Tue Feb 22 08:22:42 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1MEMgsU014918 for ; Tue, 22 Feb 2005 08:22:42 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1MEMg7u014917 for reliable_computing-outgoing; Tue, 22 Feb 2005 08:22:42 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1MEMcdu014912 for ; Tue, 22 Feb 2005 08:22:38 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1MEMbKN014910 for reliable_computing [at] interval [dot] louisiana.edu; Tue, 22 Feb 2005 08:22:37 -0600 (CST) Received: from wproxy.gmail.com (wproxy.gmail.com [64.233.184.197]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1MAQbeC014392 for ; Tue, 22 Feb 2005 04:26:43 -0600 (CST) Received: by wproxy.gmail.com with SMTP id 68so1253403wra for ; Tue, 22 Feb 2005 02:26:31 -0800 (PST) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:reply-to:to:subject:cc:mime-version:content-type:content-transfer-encoding; b=oFuH7oEb0vyG7rTIq/q6rX9aiXre9QzC3YtBUKV9wiHBpixG6si/CwNPm8KGOTG7y/hR9KPns9AHqGd336XTsc2ONYUwIuw/HrlQYL9OPLZwg6YkSS5vOjEaN+BD5SOfXYtE0u6MjagFBhaZVNZUw1ky/ftr9FeAph5HmSQ6zLs= Received: by 10.54.27.7 with SMTP id a7mr161442wra; Tue, 22 Feb 2005 02:26:31 -0800 (PST) Received: by 10.54.4.78 with HTTP; Tue, 22 Feb 2005 02:26:31 -0800 (PST) Message-ID: <81dfad1f05022202264402dab [at] mail [dot] gmail.com> Date: Tue, 22 Feb 2005 11:26:31 +0100 From: Jeff Tupper Reply-To: Jeff Tupper To: wayne [at] igor [dot] ics.uci.edu Subject: Interval solver that includes complex roots Cc: reliable_computing [at] interval [dot] louisiana.edu Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Wayne, For semialgebraic problems (those given by a collection of polynomials and quantifiers) you can also try algebraic methods (based on, e.g., cylindrical algebraic decomposition [CAD]); example programs include qepcad and qepcad b. One of the strengths of interval methods is that they can be applied to non-semialgebraic problems. Jeff From owner-reliable_computing [at] interval [dot] louisiana.edu Tue Feb 22 08:22:38 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1MEMcer014911 for ; Tue, 22 Feb 2005 08:22:38 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1MEMbUl014903 for reliable_computing-outgoing; Tue, 22 Feb 2005 08:22:37 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1MEMUGj014890 for ; Tue, 22 Feb 2005 08:22:31 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1MEMUsB014889 for reliable_computing [at] interval [dot] louisiana.edu; Tue, 22 Feb 2005 08:22:30 -0600 (CST) Received: from pickering.cc.nd.edu (pickering.cc.nd.edu [129.74.250.225]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1LJtBGt013178 for ; Mon, 21 Feb 2005 13:55:17 -0600 (CST) Received: from quartz.cheg.nd.edu (quartz.cheg.nd.edu [129.74.163.5]) by pickering.cc.nd.edu (Switch-3.1.7/Switch-3.1.7) with ESMTP id j1LJt4Ge000479 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Mon, 21 Feb 2005 14:55:06 -0500 (EST) Date: Mon, 21 Feb 2005 14:55:02 -0500 (EST) From: Youdong Lin X-X-Sender: ylin [at] quartz [dot] cheg.nd.edu To: Wayne Hayes cc: reliable_computing [at] interval [dot] louisiana.edu Subject: Re: Interval solver that includes complex roots In-Reply-To: <20050218025902.GA9230 [at] igor [dot] ics.uci.edu> Message-ID: References: <20050217051336.GA28708 [at] igor [dot] ics.uci.edu> <20050218025902.GA9230 [at] igor [dot] ics.uci.edu> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-ND-MTA-Date: Mon, 21 Feb 2005 14:55:09 -0500 (EST) X-ND-Virus-Scan: engine v4.3.20; dat v4431 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk We could let x_j = a_j + ib_j, where a_j and b_j are both real. In this way, we still can use solvers for systems of equations to find all REAL roots of the resulting 2n by 2n system, instead of the original n by n system. Youdong On Thu, 17 Feb 2005, Wayne Hayes wrote: > Let me re-phrase my question. I would like to know if there exists an > interval solver that includes finding complex roots. It would be > sufficient if the solver solved only systems of polynomial equations, > but they have to be of arbitrary degree. > > The application, surprisingly enough, is a branch of quantum mechanics > called Yang-Mills Theory, which is a branch of what's called "string > theory", related to QCD. I just saw a talk by a physicist yesterday. > > She has a problem that reduces to needing to know the number of roots > of a system of polynomial equations. She doesn't even need to know the > actual value of the roots, just how many there are. Furthermore, even > a lower bound on how many roots there are would be a good start; we > don't absolutely need to resolve close roots, for example, although > of course it would be nice. The equations are of the form > > \lambda_i = C_i \sum_{k=0}^d a_k (x_i)^k, i=1,...,n > 0 = \sum_{i=1}^n C_i l_i (x_i)^k , k=0,...,d > > and we want to solve for the x_i's. She's tried Mathematica > and I think Maple, and I told her that I knew in principle > that an interval solver (similar to GrafEq) should be able to > solve this rigorously -- but of course it may turn out to > scale exponentially in d or n. However, we cannot ignore > complex roots, as they are crucial in computing quantum > mechanical wave functions. > > I would guess that it would, in principle, not be a major > step to take some existing solver and extend it to include > complex roots. The question is whether it's practical. > > - Wayne > > From owner-reliable_computing [at] interval [dot] louisiana.edu Tue Feb 22 09:09:56 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1MF9t8Y015028 for ; Tue, 22 Feb 2005 09:09:56 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1MF9tB8015027 for reliable_computing-outgoing; Tue, 22 Feb 2005 09:09:55 -0600 (CST) Received: from marnier.ucs.louisiana.edu (root [at] marnier [dot] ucs.louisiana.edu [130.70.132.233]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1MF9lmc015023 for ; Tue, 22 Feb 2005 09:09:52 -0600 (CST) Received: from Liberty (h158065.louisiana.edu [130.70.158.65]) by marnier.ucs.louisiana.edu (8.13.1/8.13.1/ull-ucs-mx-host_1.9) with SMTP id j1MF9f14021500; Tue, 22 Feb 2005 09:09:46 -0600 (CST) Message-Id: <2.2.32.20050222150940.00a25fec [at] pop [dot] louisiana.edu> X-Sender: rbk5287 [at] pop [dot] louisiana.edu X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Tue, 22 Feb 2005 09:09:40 -0600 To: Wayne Hayes , reliable_computing [at] interval [dot] louisiana.edu From: "R. Baker Kearfott" Subject: Re: Interval solver that includes complex roots Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Wayne (et al), If you can send me a sample problem, I might be able to code it up for you for GlobSol as an example and to see how it works (provided coding it is not a major project). Best regards, Baker At 12:30 PM 2/21/2005 -0800, Wayne Hayes wrote: >On Mon, Feb 21, 2005 at 02:55:02PM -0500, Youdong Lin wrote: >> We could let x_j = a_j + ib_j, where a_j and b_j are both real. >> In this way, we still can use solvers for systems of >> equations to find all REAL roots of the resulting 2n by 2n system, >> instead of the original n by n system. > >Yes, this is true and is easy to do, assuming we have a solver than can >solve a 2n by 2n system. Does such a solver currently exist? I know >it can be done and I know how to do it, but it would be alot of work >to actually code it up. Certainly this must have been done already? > >GrafEq would be perfect except it only works in two variables, whereas >I need to solve systems of polynomial equations with an arbitrary >number of equations and variables. > >UniCalc appears to have the language to define the problem nicely, but >is not rigorous and returns answers which are only approximate, and >will often claim to find multiple nearby roots where in fact only one >actually exists. > >INTLAB looks like it might be up to the task, but I am not very >famaliar to MATLAB and am reluctant to expend the effort to learn it >unless I know there's a decent chance I can get it to work. > >GlobSol also looks promising. > >There are plenty of packages listed at > > http://www.cs.utep.edu/interval-comp/intsoft.html > >but I just want some hints as to were to start! :-) > > - Wayne > > --------------------------------------------------------------- R. Baker Kearfott, rbk [at] louisiana [dot] edu (337) 482-5346 (fax) (337) 482-5270 (work) (337) 993-1827 (home) URL: http://interval.louisiana.edu/kearfott.html Department of Mathematics, University of Louisiana at Lafayette (Room 217 Maxim D. Doucet Hall, 1403 Johnston Street) Box 4-1010, Lafayette, LA 70504-1010, USA --------------------------------------------------------------- From owner-reliable_computing [at] interval [dot] louisiana.edu Tue Feb 22 09:48:57 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1MFmvXi015099 for ; Tue, 22 Feb 2005 09:48:57 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1MFmuUP015098 for reliable_computing-outgoing; Tue, 22 Feb 2005 09:48:56 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1MFmmLK015094 for ; Tue, 22 Feb 2005 09:48:53 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j1MFmh606217; Tue, 22 Feb 2005 08:48:43 -0700 (MST) Message-Id: <200502221548.j1MFmh606217 [at] cs [dot] utep.edu> Date: Tue, 22 Feb 2005 08:48:41 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: ISIPTA'05 last CFP To: reliable_computing [at] interval [dot] louisiana.edu, interval [at] cs [dot] utep.edu, nafips-l [at] mailbox [dot] gsu.edu Cc: fgcozman [at] usp [dot] br MIME-Version: 1.0 Content-Type: TEXT/plain; charset=ISO-8859-1 Content-MD5: sguYC16Obl//QbW0H4Qn0Q== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from QUOTED-PRINTABLE to 8bit by interval.louisiana.edu id j1MFmsLK015095 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk forwarding. ------------- Begin Forwarded Message ------------- From: "ISIPTA '05" Help circulating this announcement is appreciated; sorry for duplicates. Please note: ISIPTA'05: final deadline extension, increase in paper size (details in final call for papers below) +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ ISIPTA '05 4th International Symposium on Imprecise Probabilities and Their Applications Final Call for Papers July 20-23, 2005 Carnegie Mellon University Pittsburgh, Pennsylvania, USA http://www.sipta.org/isipta05 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ The ISIPTA meetings are one of the primary international forums to present and discuss new results on the theory and applications of imprecise probabilities. Imprecise probability has a wide scope, being a generic term for the many mathematical or statistical models which measure chance or uncertainty without sharp numerical probabilities. These models include belief functions, Choquet capacities, comparative probability orderings, convex sets of probability measures, fuzzy measures, interval-valued probabilities, possibility measures, plausibility measures, and upper and lower expectations or previsions. Imprecise probability models are needed in inference problems where the relevant information is scarce, vague or conflicting, and in decision problems where preferences may also be incomplete. Themes of the symposium ----------------------- Although the symposium will be open to contributions on all aspects of imprecise probability, three main themes will be emphasised: decision-making, algorithms, and real applications. Topics of interest include, but are not limited to: models of coherent imprecise assessments convex sets of probability measures (credal sets) interval-valued probabilities upper and lower expectations or previsions non-additive set functions, and in particular Choquet capacities (and Choquet integration), fuzzy measures, possibility measures, belief and plausibility measures random sets rough sets comparative probability orderings qualitative reasoning about uncertainty imprecision in utilities and expected utilities limit laws for imprecise probabilities physical models of imprecise probability philosophical foundations for imprecise probabilities psychological models for imprecision and indeterminacy in probability assessments elicitation techniques for imprecise probabilities robust statistics probabilistic bounding analysis data mining with imprecise probabilities/missing data estimation and learning of imprecise probability models decision making with imprecise probabilities ambiguity aversion and economic models of imprecise probability uncertainty in financial markets algorithms for manipulating imprecise probabilities Dempster-Shafer theory information algebras and probabilistic argumentation systems probabilistic logic, propositional and first-order credal networks and other graphical models credal classification applications in statistics, economics, finance, management, engineering, computer science and artificial intelligence, psychology, philosophy and related fields Workshop on Financial Risk Assessment ------------------------------------- There will be a workshop addendum to the conference, to be held on July 24, with invited speakers on the topic of financial risk assessment, to which all of the ISIPTA'05 participants are welcome, at no additional registration cost. Details will be announced later. Location -------- ISIPTA '05 will be held at Carnegie Mellon University, in Pittsburgh, Pennsylvania, United States. Important dates --------------- (These dates reflect the last and final deadline extension!) Paper submission deadline: March 7 2005 (11:59PM US Eastern time) Notification of acceptance: April 25 2005 Deadline for revised papers: May 25 2005 Symposium: July 20-23 2005 Submissions and Proceedings --------------------------- (Note increase in paper size!!) We will accept papers only as PDF files with a limit of 10 pages in two-column format. Guidelines are available both in a PDF file and in a word file at the site http://www.sipta.org/isipta05/submit.html. The simplest way to produce an article that satisfies these requirements is to use Latex with the isipta2005.sty style, using guidelinesat the file isipta05.tex (or using this file as a template). Otherwise, use the file isipta2005.doc as template. Submission is electronic. You should introduce paper data and attach the PDF file by using the submission page. The Program Committee will decide which of the submitted papers are accepted, by carefully evaluating their originality, significance, technical soundness, and clarity of exposition. All the accepted papers will be included in a volume of proceedings, published by Brightdocs. There will be free electronic access to the proceedings after some time. Before the conference, the electronic access will be restricted to the ISIPTA '05 attendees to allow them to study the papers in some detail before they are actually presented. Each accepted paper will be given the opportunity for both a brief oral presentation as well as a poster session. Program Board ------------- Fabio Cozman (Universidade de Sao Paulo, Brazil) Robert Nau (Duke University, USA) Teddy Seidenfeld (Carnegie Mellon University, USA) Steering Committee ------------------ Gert de Cooman (Universiteit Gent, Belgium) Fabio G. Cozman (Universidade de São Paulo, Brazil) Serafin Moral (Universidad de Granada, Spain) Robert Nau (Duke University, USA) Teddy Seidenfeld (Carnegie Mellon University, USA) Marco Zaffalon (IDSIA-Istituto Dalle Molle di Studi sull'Intelligenza Artificiale, Switzerland) The members of the program committee are listed in the site www.sipta.org/isipta05.html. Questions --------- If you have any questions about the symposium, please contact the Organising Committee preferably by email (teddy [at] stat [dot] cmu.edu - fgcozman [at] usp [dot] br), or at the following address: Teddy Seidenfeld Department of Statistics Carnegie Mellon University Pittsburgh PA 15213 Phone: 412 - 268 - 2209 Fax: 412 - 268 - 1440 ------------- End Forwarded Message ------------- From owner-reliable_computing [at] interval [dot] louisiana.edu Thu Feb 24 06:07:29 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1OC7SdY019211 for ; Thu, 24 Feb 2005 06:07:29 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1OC7SMO019210 for reliable_computing-outgoing; Thu, 24 Feb 2005 06:07:28 -0600 (CST) Received: from smtp2.rz.tu-harburg.de (smtp2.rz.tu-harburg.de [134.28.205.13]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1OC7JMQ019206 for ; Thu, 24 Feb 2005 06:07:25 -0600 (CST) Received: from mail2.rz.tu-harburg.de (mail2.rz.tu-harburg.de [134.28.202.179]) by smtp2.rz.tu-harburg.de (8.13.1/8.13.1) with ESMTP id j1OC6bNX002146 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK); Thu, 24 Feb 2005 13:06:37 +0100 Received: from oem-9n90y5vdt8b.net.waseda.ac.jp ([133.9.117.168]) (user=ti3sr mech=LOGIN bits=0) by mail2.rz.tu-harburg.de (8.13.1/8.13.1) with ESMTP id j1OC6Ujl025156 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NO); Thu, 24 Feb 2005 13:06:35 +0100 Date: Thu, 24 Feb 2005 13:06:27 -0100 To: "Sergey P. Shary" , reliable_computing [at] interval [dot] louisiana.edu Subject: Re: INTLAB problems References: <003001c51591$4c979190$7c00a8c0@COMP0> <001b01c51a3f$d7cff880$7c00a8c0@COMP0> From: "Siegfried M. Rump" Organization: Technical University Hamburg-Harburg Content-Type: text/plain; format=flowed; delsp=yes; charset=utf-8 MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Message-ID: In-Reply-To: <001b01c51a3f$d7cff880$7c00a8c0@COMP0> User-Agent: Opera M2/7.54 (Win32, build 3869) X-Scanned-By: TUHH Rechenzentrum content checker on 134.28.205.13 X-Scanned-By: TUHH Rechenzentrum content checker on 134.28.202.179 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Dear Sergey, it's in the Readme file. Since you are not the first with this question, I also put it into the next FAQ file. Best wishes Siegfried On Thu, 24 Feb 2005 13:09:45 +0600, Sergey P. Shary wrote: > > > > Many thanks to all those who responded to my question on INTLAB features: > to Prof. Sigfried Rump, Mr. I.A. Ashokaraj and Dr. Takeshi Ogita. You > really > helped me with your advices. > > > The only remaining question (maybe, suggestion) is: > why isn't that > thoroughly documented? > > > > Sergey P. Shary > > > > > > > ----- Original Message ----- > From: "Siegfried M. Rump" > To: "Sergey P. Shary" ; > > Cc: "Б.С. Джаныбеков" > Sent: Monday, February 21, 2005 11:21 PM > Subject: Re: INTLAB problems > > >> Sergey, >> >> you can change that: >> >> >> intvalinit('FastIVmult'), a=[infsup(0,1) infsup(2,3)]; >> >> b=[infsup(-2,0.5) infsup(1,3)]; a*b' >> ===> Fast interval matrix multiplication in use (maximum overestimation >> factor 1.5 in radius) >> intval ans = >> [ -1.00000000000000, 10.25000000000001] >> >> >> intvalinit('SharpIVmult'), a=[infsup(0,1) infsup(2,3)]; >> >> b=[infsup(-2,0.5) infsup(1,3)]; a*b' >> ===> Slow but sharp interval matrix multiplication in use >> intval ans = >> [ 0.00000000000000, 9.50000000000000] >> >> >> >> Best wishes >> >> Siegfried >> >> >> > -- ================================================= Prof. Dr. Siegfried M. Rump Arbeitsbereich Informatik III Technische Universität Hamburg-Harburg Schwarzenbergstr. 95 21071 Hamburg Germany phone +49 40 42878 3027 secr. +49 40 42878 3227 fax +49 40 42878 2489 currently at: Faculty of Science and Engineering Waseda University 3-4-1 Okubo Shinjuku-ku Tokyo 169-8555 JAPAN From owner-reliable_computing [at] interval [dot] louisiana.edu Thu Feb 24 18:27:47 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1P0RkvG020115 for ; Thu, 24 Feb 2005 18:27:47 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1P0Rk1Q020114 for reliable_computing-outgoing; Thu, 24 Feb 2005 18:27:46 -0600 (CST) Received: from interval.louisiana.edu (rbk5287@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1P0Rh2I020110 for ; Thu, 24 Feb 2005 18:27:43 -0600 (CST) Received: (from rbk5287@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1P0RhDW020109 for reliable_computing [at] interval [dot] louisiana.edu; Thu, 24 Feb 2005 18:27:43 -0600 (CST) Received: from ns.ict.nsk.su (ns.ict.nsk.su [193.124.243.33]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1O78sjK018456 for ; Thu, 24 Feb 2005 01:09:02 -0600 (CST) Received: from COMP0 (COMP0.ict.local.net [192.168.0.124]) by ns.ict.nsk.su (8.13.1/8.13.1) with SMTP id j1O78D7R002062; Thu, 24 Feb 2005 13:08:14 +0600 (NOVT) (envelope-from shary [at] ict [dot] nsc.ru) X-Authentication-Warning: ns.ict.nsk.su: Host COMP0.ict.local.net [192.168.0.124] claimed to be COMP0 Message-ID: <001b01c51a3f$d7cff880$7c00a8c0@COMP0> From: "Sergey P. Shary" To: , "Siegfried M. Rump" Cc: , , =?utf-8?B?0JEu0KEuINCU0LbQsNC90YvQsdC10LrQvtCy?= References: <003001c51591$4c979190$7c00a8c0@COMP0> Subject: Re: INTLAB problems Date: Thu, 24 Feb 2005 13:09:45 +0600 Organization: ICT SB RAS MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 8bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1106 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Many thanks to all those who responded to my question on INTLAB features: to Prof. Sigfried Rump, Mr. I.A. Ashokaraj and Dr. Takeshi Ogita. You really helped me with your advices. The only remaining question (maybe, suggestion) is: why isn't that thoroughly documented? Sergey P. Shary ----- Original Message ----- From: "Siegfried M. Rump" To: "Sergey P. Shary" ; Cc: "Б.С. Джаныбеков" Sent: Monday, February 21, 2005 11:21 PM Subject: Re: INTLAB problems > Sergey, > > you can change that: > > >> intvalinit('FastIVmult'), a=[infsup(0,1) infsup(2,3)]; > >> b=[infsup(-2,0.5) infsup(1,3)]; a*b' > ===> Fast interval matrix multiplication in use (maximum overestimation > factor 1.5 in radius) > intval ans = > [ -1.00000000000000, 10.25000000000001] > > >> intvalinit('SharpIVmult'), a=[infsup(0,1) infsup(2,3)]; > >> b=[infsup(-2,0.5) infsup(1,3)]; a*b' > ===> Slow but sharp interval matrix multiplication in use > intval ans = > [ 0.00000000000000, 9.50000000000000] > >> > > Best wishes > > Siegfried > > > From owner-reliable_computing [at] interval [dot] louisiana.edu Fri Feb 25 05:55:49 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1PBtm3v021504 for ; Fri, 25 Feb 2005 05:55:48 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1PBtmsF021503 for reliable_computing-outgoing; Fri, 25 Feb 2005 05:55:48 -0600 (CST) Received: from GreenSrv.rz.unibw-muenchen.de (greensrv.RZ.UniBw-Muenchen.de [137.193.10.35]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1PBtcJM021499 for ; Fri, 25 Feb 2005 05:55:44 -0600 (CST) Received: from localhost (GreenSrv [127.0.0.1]) by GreenSrv.rz.unibw-muenchen.de (8.12.8/8.12.8) with ESMTP id j1PBq6Gu019514; Fri, 25 Feb 2005 12:52:06 +0100 Received: from GreenSrv.rz.unibw-muenchen.de ([127.0.0.1]) by localhost (GreenSrv [127.0.0.1]) (amavisd-new, port 10024) with LMTP id 19436-01; Fri, 25 Feb 2005 12:52:02 +0100 (CET) Received: from informatik.unibw-muenchen.de (riemann.Informatik.UniBw-Muenchen.de [137.193.61.36]) by GreenSrv.rz.unibw-muenchen.de (8.12.8/8.12.8) with ESMTP id j1PBpVTt019485 (version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NO); Fri, 25 Feb 2005 12:51:31 +0100 X-Remarks: If SPAM is relayed via GreenSrv.rz.unibw-muenchen.de to outside of unibw-muenchen.de, please report it to abuse@unibw-muenchen.de Message-ID: <421F113C.60106 [at] informatik [dot] unibw-muenchen.de> Date: Fri, 25 Feb 2005 12:51:24 +0100 From: Peter Hertling User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; de-DE; rv:1.0.1) Gecko/20020823 Netscape/7.0 X-Accept-Language: de-de, de MIME-Version: 1.0 To: cca-list@fernuni-hagen.de, COMP-THY [at] LISTSERV [dot] ND.EDU, fom [at] cs [dot] nyu.edu, comprox [at] doc [dot] ic.ac.uk, na.digest@na-net.ornl.gov, reliable_computing [at] interval [dot] louisiana.edu, THEORYNT [at] LISTSERV [dot] NODAK.EDU Subject: CCA 2005 - Second Call for Papers Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 8bit X-Virus-Scanned: by amavisd-new at GreenSrv.rz.unibw-muenchen.de Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk ________________________________________________________________ Second Announcement and Call for Papers ________________________________________________________________ C C A 2 0 0 5 Second International Conference on Computability and Complexity in Analysis August 25-29, 2005, Kyoto, Japan 25-26: Satellite seminars 27-29: Main conference http://cca-net.de/cca2005 ________________________________________________________________ Submissions: Authors are invited to submit a PostScript or PDF version of a paper to the email address provided on the web site of the conference. Dates: Submission: May 1, 2005 Notification: June 5, 2005 Camera-ready version: July 3, 2005 Satellite seminars and reception: August 25-26, 2005 Main conference: August 27-29, 2005 Scope The conference is concerned with the theory of computability and complexity over real-valued data. Computability and complexity theory are two central areas of research in mathematical logic and theoretical computer science. Computability theory is the study of the limitations and abilities of computers in principle. Computational complexity theory provides a framework for understanding the cost of solving computational problems, as measured by the requirement for resources such as time and space. The classical approach in these areas is to consider algorithms as operating on finite strings of symbols from a finite alphabet. Such strings may represent various discrete objects such as integers or algebraic expressions, but cannot represent general real or complex numbers, unless they are rounded. Most mathematical models in physics and engineering, however, are based on the real number concept. Thus, a computability theory and a complexity theory over the real numbers and over more general continuous data structures is needed. Unlike the well established classical theory over discrete structures, the theory of computation over continuous data is still in its infancy, despite remarkable progress in recent years. Many important fundamental problems have not yet been studied, and presumably numerous unexpected and surprising results are waiting to be detected. Scientists working in the area of computation on real-valued data come from different fields, such as theoretical computer science, domain theory, logic, constructive mathematics, computer arithmetic, numerical mathematics and all branches of analysis. The conference provides a unique opportunity for people from such diverse areas to meet and exchange ideas and knowledge. The topics of interest include foundational work on various models and approaches for describing computability and complexity over the real numbers. They also include complexity-theoretic investigations, both foundational and with respect to concrete problems, and new implementations of exact real arithmetic, as well as further developments of already existing software packages. We hope to gain new insights into computability-theoretic aspects of various computational questions from physics and from other fields involving computations over the real numbers. Scientific Program Committee: Vasco Brattka (Cape Town, South Africa) Peter Hertling, chair (Munich, Germany) Hajime Ishihara (Ishikawa, Japan) Iraj Kalantari (Macomb, USA) Ker-I Ko (Stony Brook, USA) Vladik Kreinovich (El Paso, USA) Jack H. Lutz (Ames, USA) Joseph S. Miller (Bloomington, USA) Robert Rettinger (Hagen, Germany) Matthias Schröder (Edinburgh, Scotland) Alex Simpson (Edinburgh, Scotland) Klaus Weihrauch (Hagen, Germany) Atsushi Yoshikawa (Kyushu, Japan) Xizhong Zheng (Cottbus, Germany) Ning Zhong (Cincinnati, USA) Martin Ziegler (Odense, Denmark) Organizing Committee: Hiroyasu Kamo (Nara, Japan) Takakazu Mori (Kyoto, Japan) Izumi Takeuti (Toho, Japan) Hideki Tsuiki, chair (Kyoto, Japan) Yoshiki Tsujii (Kyoto, Japan) Mariko Yasugi (Kyoto, Japan) Invited Speakers Vasco Brattka (Cape Town, South Africa) Masami Hagiya (Tokyo, Japan) Daisuke Takahashi (Waseda, Japan) Proceedings A technical report including the accepted papers will be distributed at the conference. It is planned to publish a special issue of the Journal of Complexity dedicated to the conference. After the workshop, the participants will be invited to submit their papers for publication in this special issue. The papers will be subject to the usual refereeing process of the journal. Satellite Seminars On August 25 and 26, 2005, there will be satellite seminars that will consist of introductory lectures to CCA and related areas. The following speakers have agreed to give introductory lectures (topic in brackets): Andrej Bauer (Realizability for constructive and computable mathematics) Martin Escardo (Synthetic topology of computational spaces) Peter Hertling (Computable analysis via representations) Hideyuki Suzuki (Analog computation) Mariko Yasugi (Computable versions of basic theorems in functional analysis) Atsushi Yoshikawa (Computable versions of basic theorems in functional analysis) More detailed information will soon be available on the web site of the conference. Venue The conference will be held on the Yoshida Campus of Kyoto University. Kyoto was the imperial capital of Japan for over 1000 years until 1867. It has many famous temples, shrines, national treasures, and festivals, and it is still the cultural center of Japan. Kyoto is also known as an academic city with many universities and research institutes. Local Information Kyoto is about 100 km north-east of Kansai International Airport. Travel and on-site information will soon be available on the web site of the conference. Support: We expect that some financial support will be available to cover part of the travel and hotel expenses of some of the speakers. Details will be announced soon. The conference is supported by the Graduate School of Human and Environmental Studies, Kyoto University. ________________________________________________________________ From owner-reliable_computing [at] interval [dot] louisiana.edu Fri Feb 25 09:07:17 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1PF7G4d021760 for ; Fri, 25 Feb 2005 09:07:16 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1PF7Gk0021759 for reliable_computing-outgoing; Fri, 25 Feb 2005 09:07:16 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1PF77SL021755 for ; Fri, 25 Feb 2005 09:07:13 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j1PF6lE00812; Fri, 25 Feb 2005 08:06:47 -0700 (MST) Message-Id: <200502251506.j1PF6lE00812 [at] cs [dot] utep.edu> Date: Fri, 25 Feb 2005 08:06:46 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: Deadline Extended To: reliable_computing [at] interval [dot] louisiana.edu, interval [at] cs [dot] utep.edu Cc: iajwa [at] ashland [dot] edu MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: 5UkRmWG+ep5890kHomiTsw== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk forwarding P.S. I have a new computer with which I am still somewhat struggling, and our network is going to be repaired over the weekend. I apologize for possible further delays in answering emails. Vladik ------------- Begin Forwarded Message ------------- Date: Fri, 25 Feb 2005 09:08:34 -0500 From: "Iyad Ajwa" Subject: Deadline Extended Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline Dear Colleagues, Upon request from several colleagues, I am pleased to announce that the deadline to submit a poster has been extended to March 1, 2005. Please be reminded of the following: 1. The deadline to submit a poster abstract is now March 1, 2005. 2. The deadline to submit a request for travel support is March 1, 2005. 3. The deadline to register is March 1, 2005. 4. Register and/or submit the abstract of your poster online at http://www.ashland.edu/~eccad and click on Registration. 5. Travel infomration, maps, directions, travel by air, and other accommodations are posted on the conference Website at http://www.ashland.edu/~eccad and click on Accommodations or click on Travel to ECCAD. The two pages are linked to each other. For more information, please see the original CFP below. Thank you. ============================================ ECCAD 2005 The 12th Annual East Coast Computer Algebra Day CALL FOR PARTICIPATION ============================================ HOSTED BY: The Department of Mathematics and Computer Science Ashland University Ashland, Ohio, USA DATE: March 12, 2005 CONFERENCE WEBSITE: http://www.ashland.edu/~eccad CONFERNCE EMAIL: eccad [at] ashland [dot] edu ========================================= ECCAD, East Coast Computer Algebra Day, is an annual conference that provides opportunities to learn and share new developments and to present research results in the areas of symbolic and mathematical computation. The 12th Annual East Coast Computer Algebra Day (ECCAD 2005) will be held on Saturday, March 12, 2005. It will be hosted by the Department of Mathematics and Computer Science at Ashland University in Ashland, Ohio, USA. ECCAD 2005 will be in honor of Professor Bob Caviness's 65th birthday. Bob Caviness is a Professor Emeritus (with a joint appointment in Mathematical Sciences) and former Chair of the Department of Computer & Information Sciences at the University of Delaware. His research interests include the design and analysis of algebraic algorithms, in particular, algorithmic decision procedures for integration in finite terms and for computing closed form solutions of differential equations.More recently he has turned to web-based technologies. He has five Ph.D. and two M.S students. Professor Caviness has served in various editorial capacities for the following jourmals: ACM Transactions on Mathematical Software, ACM Computing Reviews, CONSTRAINTS, the online e-commerce journal ACM SIGecom Exchanges, and the Journal of Symbolic Computation. He served as Editor-in-Chief of the latter journal for the period 1995-2000. He has also lectured widely and served the profession in many other capacities. Happy Birthday, Bob! THEMES: - Algebraic Algorithms - Symbolic-Numeric Computation - Computer Algebra Systems - Mathematical Communication - Complexity of Algebraic Problems - Symbolic and Numerical Linear Algebra - Applications of Symbolic Computation INVITED SPEAKERS: The following speakers have been confirmed. Titles and abstracts will be announced on the Conference Website at http://www.ashland.edu/~eccad once they become available. 1. Professor Bruno Buchberger - Johannes Kepler University, Linz, Austria. 2. Professor Erich Kaltofen - North Carolina State University, Raleigh, NC. 3. Professor David Saunders - University of Delaware, Newark, DE. 4. Professor Paul Wang - Kent State University, kent, OH. REGISTRATION: Online Registration is available on the Conference Website at http://www.ashland.edu/~eccad. POSTER SESSIONS: Two poster sessions will be held at the conference. Participants are invited to submit their posters by the submission deadline of February 25, 2005. Poster abstracts received before the submission deadline will be included in The ECCAD'2005 Book of Abstracts which will be available at the conference. POSTER SUBMISSION: Poster abstracts may be up to 2 pages long. The submission deadline is February 25, 2005. To submit a poster, either use the Online Registration Form (available on the Conference Website at http://www.ashland.edu/~eccad) to provide the title and abstract of your poster, or send the title and abstract of your poster by email to eccad [at] ashland [dot] edu TRAVEL SUPPORT: Pending NSF support, the conference hopes to be able to provide limited travel support for some participants to attend ECCAD'2005. Graduate students and junior faculty are particularly encouraged to apply. Support may cover only partially travel expenses and lodging for up to two nights. There will be no stipends. Make sure the flight has a flight number of a US carrier. If you plan to attend and apply for support, please send your inquiry with an itemized estimate of your expenses to eccad [at] ashland [dot] edu. Please note that if your application is approved and if funding is available, you will receive your reimbursement after the conference. All applications for support are due by March 1, 2005. Support is subject to successful funding. ACCOMMODATIONS: Blocks of rooms have been reserved at two local hotels for participants of ECCAD'2005. Deatiled Information are posted on the Conference Website at http://www.ashland.edu/~eccad. ADVISORY COUNCIL: - Bruce Char - Drexel University, USA - Erich Kaltofen - North Carolina State University, USA - Ilias Kotsireas - Wilfrid Laurier University, CANADA - David Saunders - University of Delaware, USA - William Sit - The City College of the City University of New York, USA - Paul S. Wang - Kent State University, USA - Stephen Watt - University of Western Ontario, CANADA ORGANIZER: Iyad A. Ajwa Department of Mathematics and Computer Science Ashland University Ashland, Ohio 44805, USA +1 419.289.5798 iajwa [at] ashland [dot] edu SPONSORS: - Ashland University - College of Arts and Sciences - Department of Mathematics and Computer Science - National Science Foundation CONFERENCE CONTACT: Iyad A. Ajwa Department of Mathematics and Computer Science Ashland University Ashland, Ohio 44805 USA Tel: +1 419.289.5798 FAX: +1 419.289.5791 iajwa [at] ashland [dot] edu eccad [at] ashland [dot] edu OTHER INFORMATION: Previous ECCAD conference homepages can be consulted at: http://www.cis.udel.edu/~saunders/eccad/. Iyad... *********************************************************** Iyad A. Ajwa, Ph.D. Associate Professor Department of Mathematics and Computer Science Ashland University Ashland, OH 44805, USA Phone: +1 419.289.5798 FAX: +1 419.289.5791 www.ashland.edu/~iajwa/ *********************************************************** ------------- End Forwarded Message ------------- From owner-reliable_computing [at] interval [dot] louisiana.edu Sun Feb 27 09:36:28 2005 Received: from interval.louisiana.edu (daemon@localhost [127.0.0.1]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1RFaMYZ026916 for ; Sun, 27 Feb 2005 09:36:22 -0600 (CST) Received: (from daemon@localhost) by interval.louisiana.edu (8.13.1/8.13.1/Submit) id j1RFaMUe026915 for reliable_computing-outgoing; Sun, 27 Feb 2005 09:36:22 -0600 (CST) Received: from cs.utep.edu (mail.cs.utep.edu [129.108.5.3]) by interval.louisiana.edu (8.13.1/8.13.1/ull-interval-math-majordomo-1.5) with ESMTP id j1RFZm1b026909 for ; Sun, 27 Feb 2005 09:36:19 -0600 (CST) Received: from aragorn (aragorn [129.108.5.35]) by cs.utep.edu (8.11.7/8.11.7) with SMTP id j1RFZSN12502 for ; Sun, 27 Feb 2005 08:35:31 -0700 (MST) Message-Id: <200502271535.j1RFZSN12502 [at] cs [dot] utep.edu> Date: Sun, 27 Feb 2005 08:35:26 -0700 (MST) From: Vladik Kreinovich Reply-To: Vladik Kreinovich Subject: Workshop in Nice on Automatic Differentiation To: reliable_computing [at] interval [dot] louisiana.edu MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: BrrptU/7FX51wp3IS4+GDA== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.4 SunOS 5.8 sun4u sparc Sender: owner-reliable_computing [at] interval [dot] louisiana.edu Precedence: bulk Forwarding, from NA Digest ********************************************* From: Laurent Hascoet Date: Mon, 21 Feb 2005 10:06:27 +0100 Subject: Workshop in Nice on Automatic Differentiation First European Workshop on Automatic Differentiation http://www-sop.inria.fr/tropics/adNice2005/ April 14-15, 2005 Maison du Seminaire, 29, boulevard Franck Pilatte, Nice, France This 2-day workshop replaces the UK Automatic Differentiation Workshops, which took place twice a year at Cranfield University (Shrivenham) and the University of Hertfordshire. These workshops provide a forum for the presentation of theoretical developments in and applications of Automatic Differentiation (AD) and adjoint methods. The workshop is informal and presentations on subjects such as work in progress, problem areas for AD, or possible application areas, as well as completed work are welcome. We particularly encourage PhD students and those new to the field to attend and present their work. This workshop will put some emphasis on development of AD tools and perhaps other topics to be announced later. Confirmed speakers include: Shaun Forth - "Developments in the MAD package" Andreas Griewank - Title to be confirmed Mohamed Tadjoudine - "Differentiating a time-dependent CFD solver" Jean Utke - "Automatic checkpoints and adaptive reversal schemes" Local Organiser: Laurent Hascoet - INRIA Co-organisers: Bruce Christianson - University of Hertfordshire Shaun Forth - Cranfield University Martin Bucker - RWTH Aachen Administration: Dany Sergeant Fax: +33 492 38 79 55 Please visit our web-page http://www-sop.inria.fr/tropics/adNice2005/ for further details.