DATA AND SOFTWARE INTEROPERABILITY WITH GAMS: A USER PERSPECTIVE
|
|
|
- Mervyn Blankenship
- 10 years ago
- Views:
Transcription
1 Amsterdam Optimization Modeling Group LLC Erwin Kalvelagen DATA AND SOFTWARE INTEROPERABILITY WITH GAMS: A USER PERSPECTIVE
2 Modeling Languages Specialized Modeling Languages (GAMS, AMPL, ) are very good in what they do Efficient, compact representation of a model In a way that allows thinking about the complete model And that allows and encourages experiments and reformulations Through maintainable models Especially when they become really big Handle irregular and messy data well
3 API attraction: Σειρήν Many new users are seduced to program against solver API s Familiar environment, no new funny language to learn But for large, complex models this is really a step back Even when using higher level modelingfortified API s
4 Modeling Environments Environment Solver API Modeling API Modeling Language GAMS, AMPL IBM ILOG Cplex API Concert OPL MS Solver Solver Level API SFS OML Foundation GLPK API Mathprog X Some programmers love: Malloc, Linker errors, Compiler versions, Library versions, Makefiles, Windows vs. Unix, Debuggers, But for others this is time taken away from modeling
5 Also increased use of optimization in SAS (Improved OR module) MATLAB (Tomlab, Solver Interfaces) Mathematica, Maple (Numerical Routines) GAMS, AMPL, Excel (MSF, Ilog, ) R (Solver Interfaces, Automatic Differentiation)
6 Answer: Interoperability Make GAMS more attractive for programmers by allowing to use external software Make the user decide what to do in GAMS or in other environment Make data exchange as easy as possible Even for large data Safety: this is a spot where lots of things can and will go wrong
7 Flexibility Do not decide for user Data manipulation can be done in GAMS or Excel Computation can be done in GAMS or external software Allow these decisions to be made depending on the situation Skills Available software Suitability
8 E.g. Data handling Office GAMS Where to put functionality: In Date Source Environment or Modeling System AMPL OML This has also to do with procedural vs declarative and with data manipulation capabilities
9 GDX in practice It really works Binary, fast You can look at it Limitations: Cannot add records or symbols Eg: combine two gdxfiles GDX is immutable Not self contained wrtgams: Needs declarations Zero vs non-existent GAMS interface $load is dangerous Compile time vs execution time Execution time limits (each call separate gdx, cannot add set elements at execution time)
10 Example: USDA Farm database Imports 3 MDB + XLS file GDX file ± 5 million records (raw data) File Size (bytes) FARMLandWaterResourcesDraft.mdb 80,650,240 FARMResourcesProductionDraft.mdb 429,346,86 65 symbols GTAP54-NonLandIntensive.mdb 303,66,000 FIPS2&FAOCountries.xls 29,84 farm.gdx 9,8,434 farm.gdx (compressed) 52,924,664 Good, compact way to distribute and reuse large data sets (input or output) Aggregation easier in GAMS than in Access!
11 Example: USDA Reap Model Combine data from different sources
12 Conversion mdb-> gdx $onecho > cmd.txt I=FeedGrainsData.mdb X=FeedGrainsData.gdx q=select commodity from tblcommodities s=commodity q2=select attribute from tblattributes s2=attribute q3=select period from tbltimeperiods s3=period q4=select unit from tblunits s4=unit q5=select distinct(iif(isnull(isource),'blank',isource)) \ from tblfg_update where \ not isnull(year) s5=isource q6=select geo from tblgeography s6=geocode q7=select commodity,attribute,unit,iif(isnull(isource),'blank',isource),geocode,year,period,value \ from tblfg_update where \ not isnull(year) p7=feedgrains $offecho $call
13 Typical Problems NULL s Duplicate records Multivaluedtables More difficult processing: Get latest available number Difficult in SQL and in GAMS Advantage of SQL: we can repair a number of problems on the fly.
14 Data manipulation Role of data manipulation in a modeling language OML No data manipulation at all Do it in your data source environment (e.g. Excel) AMPL More extensive data manipulation facilities Powerful if fits within declarative paradigm GAMS Extensive use of data manipulation Procedural
15 Policy Evaluation Models Often have serious data handling requirements Aggregation/disaggregation Estimation/Calibration Simulation Examples: Loc= lines of code, comments excluded Model LOC LOC for equ s Polysys < 500 Impact IntegratedIW < 500
16 Sparse Data: matrix multiplication Ampl GAMS paramn := 250; set I :={..N}; parama{iin I, j in I} := if (i=j) then ; paramb{iin I, j in I} := if (i=j) then ; paramc{iin I, j in I} := sum{k in I} A[i,k]*B[k,j]; params := sum{iin I, j in I} C[i,j]; display s; end; set i/*250/; alias (i,j,k); parameter A(i,j),B(i,j),C(i,j); A(i,i) = ; B(i,i) = ; C(i,j) = sum(k, A(i,k)*B(k,j)); scalar s; s = sum((i,j),c(i,j)); display s;
17 Timings 00 0 gams-dense gams-sparse ampl-dense ampl-sparse glpk-dense glpk-sparse
18 Solving Linear Equations Solve Ax=b for x Often not a good idea to calculate A - In GAMS we can solve by specifying Ax=b as equations
19 Inverse If you really want the inverse of a matrix: i.e. solve for A - A A = I
20 Speed Up: Advanced Basis We can provide advanced basis so the calculation takes 0 Simplex iterations Inv.m(i,j) = 0; (var: basic) Inverse.m(i,j) = ; (equ: non-basic) method= method=2 n= n= n=
21 External Solver using GDX execute_unload'a.gdx',i,a; execute '=invert.exe a.gdx i a b.gdx pinva'; parameter pinva(i,j); execute_load'b.gdx',pinva; a.gdx GAMS i,a(i,j) Invert pinva(i,j) b.gdx
22 Test Matrix: Pei Matrix α α α + + α α method= method=2 method=3 n= n= n=
23 Other tools Cholesky Eigenvalue Eigenvector
24 Max Likelihood Estimation NLP solver can find optimal values: estimates But to get covariances we need: Hessian Invert this Hessian We can do this now in GAMS Klunky, but at least we can now do this GDX used several times
25 MLE Estimation Example * Data: * Number of days until the appearance of a carcinoma in 9 * rats painted with carcinogen DMBA. set i /i*i9/; table data(i,*) days censored i 43 i2 64 i3 88 i4 88 i5 90 i6 92 i7 206 i8 209 i9 23 i0 26 i 220 i2 227 i3 230 i4 234 i5 246 i6 265 i7 304 i8 26 i9 244 ; set k(i) 'not censored'; k(i)$(data(i,'censored')=0) = yes; parameter x(i); x(i) = data(i,'days'); scalars p 'number of observations' m 'number of uncensored observations' ; p = card(i); m = card(k); display p,m;
26 MLE Estimation * * estimation * scalar theta 'location parameter' /0/; variables sigma 'scale parameter' c 'shape parameter' loglik 'log likelihood' ; equation eloglike; c.lo = 0.00; sigma.lo = 0.00; eloglike.. loglik =e= m*log(c) - m*c*log(sigma) + (c-)*sum(k,log(x(k)-theta)) - sum(i,((x(i)-theta)/sigma)**c); model mle /eloglike/; solve mle maximizing loglik using nlp;
27 Get Hessian * * get hessian * option nlp=convert; $onecho > convert.opt hessian $offecho mle.optfile=; solve mle minimizing loglik using nlp; * * gams cannot add elements at runtime so we declare the necessary elements here * set dummy /e,x,x2/; parameter h(*,*,*) '-hessian'; execute_load "hessian.gdx",h; display h; set j /sigma,c/; parameter h0(j,j); h0('sigma','sigma') = h('e','x','x'); h0('c','c') = h('e','x2','x2'); h0('sigma','c') = h('e','x','x2'); h0('c','sigma') = h('e','x','x2'); display h0;
28 Invert Hessian * * invert hessian * execute_unload "h.gdx",j,h0; execute "=invert.exe h.gdx j h0 invh.gdx invh"; parameter invh(j,j); execute_load "invh.gdx",invh; display invh;
29 Normal Quantiles * * quantile of normal distribution * * find * p = 0.05 * q = probit(-p/2) scalar prob /.05/; * we don't have the inverse error function so we calculate it * using a small cns model equations e; variables probit; e.. Errorf(probit) =e= -prob/2; model inverterrorf /e/; solve inverterrorf using cns; display probit.l; * verification: *> qnorm(0.975); *[] *> Or just use.96
30 Finally: confidence intervals * * calculate standard errors and confidence intervals * parameter result(j,*); result('c','estimate') = c.l; result('sigma','estimate') = sigma.l; result(j,'stderr') = sqrt(abs(invh(j,j))); result(j,'conf lo') = result(j,'estimate') - probit.l*result(j,'stderr'); result(j,'conf up') = result(j,'estimate') + probit.l*result(j,'stderr'); display result; PARAMETER result estimate stderr conf lo conf up sigma c
31 New developments Instead of calling external programs with a GDX interface Call a user provided DLL With simplified syntax: Parameter A(i,j),B(j,i); A(i,j) = Scalar status; BridgeCall('gamslapack', 'invert', A, B, Status);
32 Behind the scenes Map GAMS data to fortran, c, Deal with calling conventions (stdcall) Specified in a small spec file file bridgelibrary.ini [bridge] id=gams bridge library lib=gamslapack lib2=gamsgsl subroutine invert(a,n,b,info) beginning of file gamslapack.ini [Library] Version= Description=GAMS interface to LAPack LibName=gamslapack DLLName=gamslapack Storage=F [invert] name=invert i=q // param = square matrix id=i2 i2=d // param2 = n o=q // param3 = square matrix od=i,2 od2=i, o2=n // param4 = info status=o2
33 Gets more exciting when We can parse subroutine headers In libraries And generate this automatically This will open up access to Numerical, statistical libraries Current tools as function (sql2gms, LS solver, ) Etc. Longer term: also for equations derivatives
Gamma Distribution Fitting
Chapter 552 Gamma Distribution Fitting Introduction This module fits the gamma probability distributions to a complete or censored set of individual or grouped data values. It outputs various statistics
An Overview Of Software For Convex Optimization. Brian Borchers Department of Mathematics New Mexico Tech Socorro, NM 87801 borchers@nmt.
An Overview Of Software For Convex Optimization Brian Borchers Department of Mathematics New Mexico Tech Socorro, NM 87801 [email protected] In fact, the great watershed in optimization isn t between linearity
Programming Languages
Programming Languages Programming languages bridge the gap between people and machines; for that matter, they also bridge the gap among people who would like to share algorithms in a way that immediately
SAS Certificate Applied Statistics and SAS Programming
SAS Certificate Applied Statistics and SAS Programming SAS Certificate Applied Statistics and Advanced SAS Programming Brigham Young University Department of Statistics offers an Applied Statistics and
Expedite for Windows Software Development Kit Programming Guide
GXS EDI Services Expedite for Windows Software Development Kit Programming Guide Version 6 Release 2 GC34-3285-02 Fifth Edition (November 2005) This edition replaces the Version 6.1 edition. Copyright
1. Overview. 2. Requirements
MDB2GMS: A TOOL FOR IMPORTING MS ACCESS DATABASE TABLES INTO GAMS ERWIN KALVELAGEN Abstract. This document describes the MDB2GMS utility which allows to convert data stored in MS Access database file into
TOMLAB - For fast and robust largescale optimization in MATLAB
The TOMLAB Optimization Environment is a powerful optimization and modeling package for solving applied optimization problems in MATLAB. TOMLAB provides a wide range of features, tools and services for
CPLEX Tutorial Handout
CPLEX Tutorial Handout What Is ILOG CPLEX? ILOG CPLEX is a tool for solving linear optimization problems, commonly referred to as Linear Programming (LP) problems, of the form: Maximize (or Minimize) c
Matrix Multiplication
Matrix Multiplication CPS343 Parallel and High Performance Computing Spring 2016 CPS343 (Parallel and HPC) Matrix Multiplication Spring 2016 1 / 32 Outline 1 Matrix operations Importance Dense and sparse
SUMAN DUVVURU STAT 567 PROJECT REPORT
SUMAN DUVVURU STAT 567 PROJECT REPORT SURVIVAL ANALYSIS OF HEROIN ADDICTS Background and introduction: Current illicit drug use among teens is continuing to increase in many countries around the world.
STATISTICA Formula Guide: Logistic Regression. Table of Contents
: Table of Contents... 1 Overview of Model... 1 Dispersion... 2 Parameterization... 3 Sigma-Restricted Model... 3 Overparameterized Model... 4 Reference Coding... 4 Model Summary (Summary Tab)... 5 Summary
Mathematical Libraries on JUQUEEN. JSC Training Course
Mitglied der Helmholtz-Gemeinschaft Mathematical Libraries on JUQUEEN JSC Training Course May 10, 2012 Outline General Informations Sequential Libraries, planned Parallel Libraries and Application Systems:
6.088 Intro to C/C++ Day 4: Object-oriented programming in C++ Eunsuk Kang and Jean Yang
6.088 Intro to C/C++ Day 4: Object-oriented programming in C++ Eunsuk Kang and Jean Yang Today s topics Why objects? Object-oriented programming (OOP) in C++ classes fields & methods objects representation
IBM ILOG CPLEX Optimization Studio Getting Started with CPLEX. Version12Release4
IBM ILOG CPLEX Optimization Studio Getting Started with CPLEX Version12Release4 Copyright notice Describes general use restrictions and trademarks related to this document and the software described in
CE 504 Computational Hydrology Computational Environments and Tools Fritz R. Fiedler
CE 504 Computational Hydrology Computational Environments and Tools Fritz R. Fiedler 1) Operating systems a) Windows b) Unix and Linux c) Macintosh 2) Data manipulation tools a) Text Editors b) Spreadsheets
Logistic Regression. Vibhav Gogate The University of Texas at Dallas. Some Slides from Carlos Guestrin, Luke Zettlemoyer and Dan Weld.
Logistic Regression Vibhav Gogate The University of Texas at Dallas Some Slides from Carlos Guestrin, Luke Zettlemoyer and Dan Weld. Generative vs. Discriminative Classifiers Want to Learn: h:x Y X features
Logistic Regression. Jia Li. Department of Statistics The Pennsylvania State University. Logistic Regression
Logistic Regression Department of Statistics The Pennsylvania State University Email: [email protected] Logistic Regression Preserve linear classification boundaries. By the Bayes rule: Ĝ(x) = arg max
Machine Learning and Pattern Recognition Logistic Regression
Machine Learning and Pattern Recognition Logistic Regression Course Lecturer:Amos J Storkey Institute for Adaptive and Neural Computation School of Informatics University of Edinburgh Crichton Street,
From the help desk: Demand system estimation
The Stata Journal (2002) 2, Number 4, pp. 403 410 From the help desk: Demand system estimation Brian P. Poi Stata Corporation Abstract. This article provides an example illustrating how to use Stata to
MATLAB Tutorial. Chapter 6. Writing and calling functions
MATLAB Tutorial Chapter 6. Writing and calling functions In this chapter we discuss how to structure a program with multiple source code files. First, an explanation of how code files work in MATLAB is
Parallel and Distributed Computing Programming Assignment 1
Parallel and Distributed Computing Programming Assignment 1 Due Monday, February 7 For programming assignment 1, you should write two C programs. One should provide an estimate of the performance of ping-pong
Jonathan Worthington Scarborough Linux User Group
Jonathan Worthington Scarborough Linux User Group Introduction What does a Virtual Machine do? Hides away the details of the hardware platform and operating system. Defines a common set of instructions.
Building Applications Using Micro Focus COBOL
Building Applications Using Micro Focus COBOL Abstract If you look through the Micro Focus COBOL documentation, you will see many different executable file types referenced: int, gnt, exe, dll and others.
15.062 Data Mining: Algorithms and Applications Matrix Math Review
.6 Data Mining: Algorithms and Applications Matrix Math Review The purpose of this document is to give a brief review of selected linear algebra concepts that will be useful for the course and to develop
Corporate Defaults and Large Macroeconomic Shocks
Corporate Defaults and Large Macroeconomic Shocks Mathias Drehmann Bank of England Andrew Patton London School of Economics and Bank of England Steffen Sorensen Bank of England The presentation expresses
Lecture 5: Singular Value Decomposition SVD (1)
EEM3L1: Numerical and Analytical Techniques Lecture 5: Singular Value Decomposition SVD (1) EE3L1, slide 1, Version 4: 25-Sep-02 Motivation for SVD (1) SVD = Singular Value Decomposition Consider the system
Comparing SQL and NOSQL databases
COSC 6397 Big Data Analytics Data Formats (II) HBase Edgar Gabriel Spring 2015 Comparing SQL and NOSQL databases Types Development History Data Storage Model SQL One type (SQL database) with minor variations
Embedded Systems. Review of ANSI C Topics. A Review of ANSI C and Considerations for Embedded C Programming. Basic features of C
Embedded Systems A Review of ANSI C and Considerations for Embedded C Programming Dr. Jeff Jackson Lecture 2-1 Review of ANSI C Topics Basic features of C C fundamentals Basic data types Expressions Selection
Computational Mathematics with Python
Computational Mathematics with Python Basics Claus Führer, Jan Erik Solem, Olivier Verdier Spring 2010 Claus Führer, Jan Erik Solem, Olivier Verdier Computational Mathematics with Python Spring 2010 1
PGR Computing Programming Skills
PGR Computing Programming Skills Dr. I. Hawke 2008 1 Introduction The purpose of computing is to do something faster, more efficiently and more reliably than you could as a human do it. One obvious point
Matrox Imaging White Paper
Vision library or vision specific IDE: Which is right for you? Abstract Commercial machine vision software is currently classified along two lines: the conventional vision library and the vision specific
How to use PDFlib products with PHP
How to use PDFlib products with PHP Last change: July 13, 2011 Latest PDFlib version covered in this document: 8.0.3 Latest version of this document available at: www.pdflib.com/developer/technical-documentation
Mathematical Libraries and Application Software on JUROPA and JUQUEEN
Mitglied der Helmholtz-Gemeinschaft Mathematical Libraries and Application Software on JUROPA and JUQUEEN JSC Training Course May 2014 I.Gutheil Outline General Informations Sequential Libraries Parallel
GLM, insurance pricing & big data: paying attention to convergence issues.
GLM, insurance pricing & big data: paying attention to convergence issues. Michaël NOACK - [email protected] Senior consultant & Manager of ADDACTIS Pricing Copyright 2014 ADDACTIS Worldwide.
Computational Mathematics with Python
Boolean Arrays Classes Computational Mathematics with Python Basics Olivier Verdier and Claus Führer 2009-03-24 Olivier Verdier and Claus Führer Computational Mathematics with Python 2009-03-24 1 / 40
SALEM COMMUNITY COLLEGE Carneys Point, New Jersey 08069 COURSE SYLLABUS COVER SHEET. Action Taken (Please Check One) New Course Initiated
SALEM COMMUNITY COLLEGE Carneys Point, New Jersey 08069 COURSE SYLLABUS COVER SHEET Course Title Course Number Department Linear Algebra Mathematics MAT-240 Action Taken (Please Check One) New Course Initiated
Response variables assume only two values, say Y j = 1 or = 0, called success and failure (spam detection, credit scoring, contracting.
Prof. Dr. J. Franke All of Statistics 1.52 Binary response variables - logistic regression Response variables assume only two values, say Y j = 1 or = 0, called success and failure (spam detection, credit
Python, C++ and SWIG
Robin Dunn Software Craftsman O Reilly Open Source Convention July 21 25, 2008 Slides available at http://wxpython.org/oscon2008/ Python & C++ Comparisons Each is a general purpose programming language,
Advantages of PML as an iseries Web Development Language
Advantages of PML as an iseries Web Development Language What is PML PML is a highly productive language created specifically to help iseries RPG programmers make the transition to web programming and
Eliminate Memory Errors and Improve Program Stability
Eliminate Memory Errors and Improve Program Stability with Intel Parallel Studio XE Can running one simple tool make a difference? Yes, in many cases. You can find errors that cause complex, intermittent
Linear Programming. March 14, 2014
Linear Programming March 1, 01 Parts of this introduction to linear programming were adapted from Chapter 9 of Introduction to Algorithms, Second Edition, by Cormen, Leiserson, Rivest and Stein [1]. 1
AMS526: Numerical Analysis I (Numerical Linear Algebra)
AMS526: Numerical Analysis I (Numerical Linear Algebra) Lecture 19: SVD revisited; Software for Linear Algebra Xiangmin Jiao Stony Brook University Xiangmin Jiao Numerical Analysis I 1 / 9 Outline 1 Computing
Package EstCRM. July 13, 2015
Version 1.4 Date 2015-7-11 Package EstCRM July 13, 2015 Title Calibrating Parameters for the Samejima's Continuous IRT Model Author Cengiz Zopluoglu Maintainer Cengiz Zopluoglu
How to speed-up hard problem resolution using GLPK?
How to speed-up hard problem resolution using GLPK? Onfroy B. & Cohen N. September 27, 2010 Contents 1 Introduction 2 1.1 What is GLPK?.......................................... 2 1.2 GLPK, the best one?.......................................
IBM WebSphere ILOG Rules for.net
Automate business decisions and accelerate time-to-market IBM WebSphere ILOG Rules for.net Business rule management for Microsoft.NET and SOA environments Highlights Complete BRMS for.net Integration with
6. Cholesky factorization
6. Cholesky factorization EE103 (Fall 2011-12) triangular matrices forward and backward substitution the Cholesky factorization solving Ax = b with A positive definite inverse of a positive definite matrix
Mimer SQL. Programmer s Manual. Version 8.2 Copyright 2000 Mimer Information Technology AB
Mimer SQL Version 8.2 Copyright 2000 Mimer Information Technology AB Second revised edition December, 2000 Copyright 2000 Mimer Information Technology AB. Published by Mimer Information Technology AB,
Imputing Missing Data using SAS
ABSTRACT Paper 3295-2015 Imputing Missing Data using SAS Christopher Yim, California Polytechnic State University, San Luis Obispo Missing data is an unfortunate reality of statistics. However, there are
Statistics in Retail Finance. Chapter 6: Behavioural models
Statistics in Retail Finance 1 Overview > So far we have focussed mainly on application scorecards. In this chapter we shall look at behavioural models. We shall cover the following topics:- Behavioural
Javadoc like technical documentation for CAPRI
Javadoc like technical documentation for CAPRI Introduction and background - by Wolfgang Britz, July 2008 - Since 1996, CAPRI has grown to a rather complex (bio-)economic modelling system. Its code based
PEST - Beyond Basic Model Calibration. Presented by Jon Traum
PEST - Beyond Basic Model Calibration Presented by Jon Traum Purpose of Presentation Present advance techniques available in PEST for model calibration High level overview Inspire more people to use PEST!
Algorithm & Flowchart & Pseudo code. Staff Incharge: S.Sasirekha
Algorithm & Flowchart & Pseudo code Staff Incharge: S.Sasirekha Computer Programming and Languages Computers work on a set of instructions called computer program, which clearly specify the ways to carry
Computational Mathematics with Python
Numerical Analysis, Lund University, 2011 1 Computational Mathematics with Python Chapter 1: Basics Numerical Analysis, Lund University Claus Führer, Jan Erik Solem, Olivier Verdier, Tony Stillfjord Spring
Multivariate Logistic Regression
1 Multivariate Logistic Regression As in univariate logistic regression, let π(x) represent the probability of an event that depends on p covariates or independent variables. Then, using an inv.logit formulation
Similarity and Diagonalization. Similar Matrices
MATH022 Linear Algebra Brief lecture notes 48 Similarity and Diagonalization Similar Matrices Let A and B be n n matrices. We say that A is similar to B if there is an invertible n n matrix P such that
Introduction to Software Paradigms & Procedural Programming Paradigm
Introduction & Procedural Programming Sample Courseware Introduction to Software Paradigms & Procedural Programming Paradigm This Lesson introduces main terminology to be used in the whole course. Thus,
Design-Simulation-Optimization Package for a Generic 6-DOF Manipulator with a Spherical Wrist
Design-Simulation-Optimization Package for a Generic 6-DOF Manipulator with a Spherical Wrist MHER GRIGORIAN, TAREK SOBH Department of Computer Science and Engineering, U. of Bridgeport, USA ABSTRACT Robot
Chapter 6: Sensitivity Analysis
Chapter 6: Sensitivity Analysis Suppose that you have just completed a linear programming solution which will have a major impact on your company, such as determining how much to increase the overall production
ESTIMATING AVERAGE TREATMENT EFFECTS: IV AND CONTROL FUNCTIONS, II Jeff Wooldridge Michigan State University BGSE/IZA Course in Microeconometrics
ESTIMATING AVERAGE TREATMENT EFFECTS: IV AND CONTROL FUNCTIONS, II Jeff Wooldridge Michigan State University BGSE/IZA Course in Microeconometrics July 2009 1. Quantile Treatment Effects 2. Control Functions
Similar matrices and Jordan form
Similar matrices and Jordan form We ve nearly covered the entire heart of linear algebra once we ve finished singular value decompositions we ll have seen all the most central topics. A T A is positive
HPC Wales Skills Academy Course Catalogue 2015
HPC Wales Skills Academy Course Catalogue 2015 Overview The HPC Wales Skills Academy provides a variety of courses and workshops aimed at building skills in High Performance Computing (HPC). Our courses
Big-data Analytics: Challenges and Opportunities
Big-data Analytics: Challenges and Opportunities Chih-Jen Lin Department of Computer Science National Taiwan University Talk at 台 灣 資 料 科 學 愛 好 者 年 會, August 30, 2014 Chih-Jen Lin (National Taiwan Univ.)
Satisfying business needs while maintaining the
Component-Based Development With MQSeries Workflow By Michael S. Pallos Client Application Satisfying business needs while maintaining the flexibility to incorporate new requirements in a timely fashion
Distributionally Robust Optimization with ROME (part 2)
Distributionally Robust Optimization with ROME (part 2) Joel Goh Melvyn Sim Department of Decision Sciences NUS Business School, Singapore 18 Jun 2009 NUS Business School Guest Lecture J. Goh, M. Sim (NUS)
Example: Credit card default, we may be more interested in predicting the probabilty of a default than classifying individuals as default or not.
Statistical Learning: Chapter 4 Classification 4.1 Introduction Supervised learning with a categorical (Qualitative) response Notation: - Feature vector X, - qualitative response Y, taking values in C
From the help desk: Swamy s random-coefficients model
The Stata Journal (2003) 3, Number 3, pp. 302 308 From the help desk: Swamy s random-coefficients model Brian P. Poi Stata Corporation Abstract. This article discusses the Swamy (1970) random-coefficients
Typical Linear Equation Set and Corresponding Matrices
EWE: Engineering With Excel Larsen Page 1 4. Matrix Operations in Excel. Matrix Manipulations: Vectors, Matrices, and Arrays. How Excel Handles Matrix Math. Basic Matrix Operations. Solving Systems of
SQL Server Array Library 2010-11 László Dobos, Alexander S. Szalay
SQL Server Array Library 2010-11 László Dobos, Alexander S. Szalay The Johns Hopkins University, Department of Physics and Astronomy Eötvös University, Department of Physics of Complex Systems http://voservices.net/sqlarray,
Parameter Estimation for Bingham Models
Dr. Volker Schulz, Dmitriy Logashenko Parameter Estimation for Bingham Models supported by BMBF Parameter Estimation for Bingham Models Industrial application of ceramic pastes Material laws for Bingham
Standard errors of marginal effects in the heteroskedastic probit model
Standard errors of marginal effects in the heteroskedastic probit model Thomas Cornelißen Discussion Paper No. 320 August 2005 ISSN: 0949 9962 Abstract In non-linear regression models, such as the heteroskedastic
Lecture 3: Linear methods for classification
Lecture 3: Linear methods for classification Rafael A. Irizarry and Hector Corrada Bravo February, 2010 Today we describe four specific algorithms useful for classification problems: linear regression,
Mixed Integer Linear Programming in R
Mixed Integer Linear Programming in R Stefan Theussl Department of Statistics and Mathematics Wirtschaftsuniversität Wien July 1, 2008 Outline Introduction Linear Programming Quadratic Programming Mixed
Session 15 OF, Unpacking the Actuary's Technical Toolkit. Moderator: Albert Jeffrey Moore, ASA, MAAA
Session 15 OF, Unpacking the Actuary's Technical Toolkit Moderator: Albert Jeffrey Moore, ASA, MAAA Presenters: Melissa Boudreau, FCAS Albert Jeffrey Moore, ASA, MAAA Christopher Kenneth Peek Yonasan Schwartz,
Sample Size Calculation for Longitudinal Studies
Sample Size Calculation for Longitudinal Studies Phil Schumm Department of Health Studies University of Chicago August 23, 2004 (Supported by National Institute on Aging grant P01 AG18911-01A1) Introduction
InterBase 6. API Guide. Borland/INPRISE. 100 Enterprise Way, Scotts Valley, CA 95066 http://www.interbase.com
InterBase 6 API Guide Borland/INPRISE 100 Enterprise Way, Scotts Valley, CA 95066 http://www.interbase.com Inprise/Borland may have patents and/or pending patent applications covering subject matter in
CHAPTER 9 EXAMPLES: MULTILEVEL MODELING WITH COMPLEX SURVEY DATA
Examples: Multilevel Modeling With Complex Survey Data CHAPTER 9 EXAMPLES: MULTILEVEL MODELING WITH COMPLEX SURVEY DATA Complex survey data refers to data obtained by stratification, cluster sampling and/or
OpenACC 2.0 and the PGI Accelerator Compilers
OpenACC 2.0 and the PGI Accelerator Compilers Michael Wolfe The Portland Group [email protected] This presentation discusses the additions made to the OpenACC API in Version 2.0. I will also present
State of Stress at Point
State of Stress at Point Einstein Notation The basic idea of Einstein notation is that a covector and a vector can form a scalar: This is typically written as an explicit sum: According to this convention,
The C Programming Language course syllabus associate level
TECHNOLOGIES The C Programming Language course syllabus associate level Course description The course fully covers the basics of programming in the C programming language and demonstrates fundamental programming
Maximum Likelihood Estimation of Logistic Regression Models: Theory and Implementation
Maximum Likelihood Estimation of Logistic Regression Models: Theory and Implementation Abstract This article presents an overview of the logistic regression model for dependent variables having two or
Least-Squares Intersection of Lines
Least-Squares Intersection of Lines Johannes Traa - UIUC 2013 This write-up derives the least-squares solution for the intersection of lines. In the general case, a set of lines will not intersect at a
Getting off the ground when creating an RVM test-bench
Getting off the ground when creating an RVM test-bench Rich Musacchio, Ning Guo Paradigm Works [email protected],[email protected] ABSTRACT RVM compliant environments provide
An Oracle White Paper October 2013. Oracle Data Integrator 12c New Features Overview
An Oracle White Paper October 2013 Oracle Data Integrator 12c Disclaimer This document is for informational purposes. It is not a commitment to deliver any material, code, or functionality, and should
Survey, Statistics and Psychometrics Core Research Facility University of Nebraska-Lincoln. Log-Rank Test for More Than Two Groups
Survey, Statistics and Psychometrics Core Research Facility University of Nebraska-Lincoln Log-Rank Test for More Than Two Groups Prepared by Harlan Sayles (SRAM) Revised by Julia Soulakova (Statistics)
HURDLE AND SELECTION MODELS Jeff Wooldridge Michigan State University BGSE/IZA Course in Microeconometrics July 2009
HURDLE AND SELECTION MODELS Jeff Wooldridge Michigan State University BGSE/IZA Course in Microeconometrics July 2009 1. Introduction 2. A General Formulation 3. Truncated Normal Hurdle Model 4. Lognormal
Programming Languages & Tools
4 Programming Languages & Tools Almost any programming language one is familiar with can be used for computational work (despite the fact that some people believe strongly that their own favorite programming
JMulTi/JStatCom - A Data Analysis Toolkit for End-users and Developers
JMulTi/JStatCom - A Data Analysis Toolkit for End-users and Developers Technology White Paper JStatCom Engineering, www.jstatcom.com by Markus Krätzig, June 4, 2007 Abstract JStatCom is a software framework
CS3220 Lecture Notes: QR factorization and orthogonal transformations
CS3220 Lecture Notes: QR factorization and orthogonal transformations Steve Marschner Cornell University 11 March 2009 In this lecture I ll talk about orthogonal matrices and their properties, discuss
Load Manager Administrator s Guide For other guides in this document set, go to the Document Center
Load Manager Administrator s Guide For other guides in this document set, go to the Document Center Load Manager for Citrix Presentation Server Citrix Presentation Server 4.5 for Windows Citrix Access
Product Information CANape Option Simulink XCP Server
Product Information CANape Option Simulink XCP Server Table of Contents 1 Overview... 3 1.1 Introduction... 3 1.2 Overview of Advantages... 3 1.3 Application Areas... 3 1.4 Further Information... 4 2 Functions...
APPENDIX E THE ASSESSMENT PHASE OF THE DATA LIFE CYCLE
APPENDIX E THE ASSESSMENT PHASE OF THE DATA LIFE CYCLE The assessment phase of the Data Life Cycle includes verification and validation of the survey data and assessment of quality of the data. Data verification
2. Advance Certificate Course in Information Technology
Introduction: 2. Advance Certificate Course in Information Technology In the modern world, information is power. Acquiring information, storing, updating, processing, sharing, distributing etc. are essentials
Temperature Control Loop Analyzer (TeCLA) Software
Temperature Control Loop Analyzer (TeCLA) Software F. Burzagli - S. De Palo - G. Santangelo (Alenia Spazio) [email protected] Foreword A typical feature of an active loop thermal system is to guarantee
Regression III: Advanced Methods
Lecture 4: Transformations Regression III: Advanced Methods William G. Jacoby Michigan State University Goals of the lecture The Ladder of Roots and Powers Changing the shape of distributions Transforming
Dynamical Systems Analysis II: Evaluating Stability, Eigenvalues
Dynamical Systems Analysis II: Evaluating Stability, Eigenvalues By Peter Woolf [email protected]) University of Michigan Michigan Chemical Process Dynamics and Controls Open Textbook version 1.0 Creative
