Search

1 | INTRODUCTION The Joint United Nations Programme on HIV/AIDS (UNAIDS)-supported Spectrum software package is used by most countries in the world to generate annual estimates of HIV burden and [...], Introduction: The Case Surveillance and Vital Registration (CSAVR) model within Spectrum estimates HIV incidence trends from surveillance data on numbers of new HIV diagnoses and HIV-related deaths. This article describes developments of the CSAVR tool to more flexibly model diagnosis rates over time, estimate incidence patterns by sex and age group and by key population group. Methods: We modelled HIV diagnosis rate trends as a mixture of three factors, including temporal and opportunistic infection components. The tool was expanded to estimate incidence rate ratios by sex and age for countries with disaggregated reporting of new HIV diagnoses and AIDS deaths, and to account for information on key populations such as men who have sex with men (MSM), males who inject drugs (MWID), female sex workers (FSW) and females who inject drugs (FWID). We used a Bayesian framework to calibrate the tool in 71 high-income or low-HIV burden countries. Results: Across countries, an estimated median 89% (interquartile range [IQR]: 7896-9696) of HIV-positive adults knew their status in 2019. Mean CD4 counts at diagnosis were stable over time, with a median of 456 cell/[micro]l (IQR: 391-508) across countries in 2019. In European countries reporting new HIV diagnoses among key populations, median estimated proportions of males that are MSM and MWID was 1.3% (IQR: 0.996-2.0%) and 0.5696 (IQR: 0.5196-0.6496), respectively. The median estimated proportions of females that are FSW and FWID were 0.3696 (IQR: 0.2796-0.4596) and 0.14 (IQR: 0.1396-0.1596), respectively. HIV incidence per 100 person-years increased among MSM, with median estimates reaching 0.43 (IQR: 0.29-1.73) in 2019, but remained stable in MWID, FSW and FWID, at around 0.12 (IQR: 0.04-1.9), 0.09 (IQR: 0.06-0.69) and 0.1396 (IQR: 0.0896-0.9196) in 2019, respectively. Knowledge of HIV status among HIV-positive adults gradually increased since the early 1990s to exceed 7596 in more than 7596 of countries in 2019 among each key population. Conclusions: CSAVR offers an approach to using routine surveillance and vital registration data to estimate and project trends in both HIV incidence and knowledge of HIV status. Keywords: HIV case-based surveillance; HIV incidence; HIV-related mortality; key population; knowledge of status; statistical models

Math fact fluency is foundational for later mathematics education. Unfortunately, many students across the nation continue to struggle with these core skills. Computer-assisted instruction may be a potentially valuable tool for improving math fact fluency due to its ability to differentiate instruction at the student level, provide added practice opportunities, and improve student interest and motivation. However, research is currently lacking to demonstrate the effectiveness of many computer-assisted interventions. One such program is Timez Attack by Imagine Math Facts, a multiplication fact fluency training program for elementary-age students. Using a multiple baseline across groups design, we sought to determine the effectiveness of Timez Attack in improving math fact fluency in third-grade students. We randomly assigned 63 students to three study groups and regularly assessed for multiplication fact fluency for 12 weeks. Compared to baseline averages, all three study groups demonstrated improved multiplication fact fluency following the onset of the intervention phase. Further, performance during a follow-up maintenance phase demonstrated persistence of learning. The results of this study suggest that Timez Attack may be an effective computer-assisted instruction option for improving multiplication fact fluency in elementary-age students., Author(s): Andrew N. Berrett [sup.1] , Nari J. Carter [sup.1] Author Affiliations: (Aff1) Imagine Learning, , 382 W. Park Cir., 84604, Provo, UT, USA Introduction Historically, students in the USA [...]

There are no sure things in an uncertain world, as the Seattle Seahawks dramatically demonstrated to the heavily favored New Orleans Saints during the recent NFL playoffs. At about the [...]

Anasazi is a package within the Trilinos software project that provides a framework for the iterative, numerical solution of large-scale eigenvalue problems. Anasazi is written in ANSI C++ and exploits modern software paradigms to enable the research and development of eigensolver algorithms. Furthermore, Anasazi provides implementations for some of the most recent eigensolver methods. The purpose of our article is to describe the design and development of the Anasazi framework. A performance comparison of Anasazi and the popular FORTRAN 77 code ARPACK is given. Categories and Subject Descriptors: G.1.3 [Numerical Analysis]: Numerical Linear Algebra; G.4 [Mathematical Software]; D.2.13 [Software Engineering]: Reusable Software General Terms: Algorithms, Design, Performance, Reliability, Theory Additional Key Words and Phrases: Eigenvalue problems, numerical algorithms, generic programming, object-oriented programming, large-scale scientific computing DOI 10.1145/1527286.1527287

Confidence limits for variance components in mixed effects models can be readily estimated by procedures such as PROC MIXED (SAS[R]), MIXED (SPSS[R]), LME (S-PLUS[R]), and XTMIXED (Stata[R]). For a small sample unbalanced one-way random effects model case study, the REML Wald type confidence limits for the between-group variance obtained from PROC MIXED differ substantially from those from MIXED, LME, and XTMIXED. Simulations indicate that the Satterth-waite approximation used by PROC MIXED results in highly inflated confidence interval lengths, while the interval lengths from the other procedures are usually similar but also biased upward. The coverage rates of the confidence limits from all four procedures appear above the nominal value. Extreme upper limits and problems with computing the confidence limits are also possible, especially with PROC MIXED. In general, one should avoid the naive use of statistical software packages when estimating REML confidence limits for variance components with small datasets, and simulations can assist in examining the validity of the results. KEY WORDS: LME; MIXED; PROC MIXED; REML; Simulation; Variance component; XTMIXED.

Problem statement: The power of Artificial Intelligent (AI) becomes more authoritative when the system is programmed to cater the need of complex applications. MATLAB 2007B, integrating artificial intelligent system and Graphical User Interface (GUI) has reduced researchers' fear-to-model factor due to unfamiliarity and phobia to produce program codes. Approach: In this study, how GUI was developed on Matlab to model laser machining process using Adaptive Network-based Fuzzy Inference System (ANFIS) was presented. Laser cutting machine is widely known for having the most number of controllable parameters among the advanced machine tools, hence become more difficult to engineer the process into desired responses; surface roughness and kerf width. Mastering both laser processing and ANFIS programming are difficult task for most researchers, especially for the difficult to model processes. Therefore, a new approach was ventured, where GUI was developed using MATLAB integrating ANFIS variables to model the laser processing phenomenon, in which the numeric and graphical output can be easily printed to interpret the results. Results: To investigate ANFIS variables' characteristic and effect, error was analyzed via Root Mean Square Error (RMSE) and Average Percentage Error (APE). The RMSE values were then compared among various trained variables and settings to finalize best ANFIS predictive model. The results found was very promising and proved that, even a person with shallow knowledge in both artificial intelligence and laser processing can actually train the experimental data sets loaded into GUI, test and optimize ANFIS variables to make comparative analysis. Conclusion: The details of modeled work with prediction accuracy according to variable combinations were premeditated on another paper. The findings were expected to benefit precision machining industries in reducing their down time and cost as compared to the traditional way of trial and error method. Key words: ANFIS modeling, laser machining, GUI development, surface finish, kerf width, INTRODUCTION Laser cutting machine is one of new types in advanced machine tools that have been developed to improve machining process. However, when the machining process becomes more complex and [...]

In this work we propose a new algorithm to evaluate the basis functions of the Argyris finite element and their derivatives. The main novelty here is an efficient way to calculate the matrix which gives the change of coordinates between the bases of the Argyis element for the reference and for an arbitrary triangle. This matrix is factored as the product of two rectangular matrices with a strong block structure which makes their computation very easy. We show and comment on an implementation of this algorithm in Matlab. Two numerical experiments, an interpolation of a smooth function on a triangle and the finite-element solution of the Dirichlet problem for the biLaplacian, are presented in the last section to check the performance of our implementation. Categories and Subject Descriptors: G.4 [Mathematics of Computing]: Mathematics Software---Documentation; G.1.8 [Numerical Analysis]: Partial Differential Equations Finite element methods General Terms: Algorithms, Documentation Additional Key Words and Phrases: Argyris element, finite elements, Matlab DOI = 10.1145/ 1377612.1377620. http://10.1145/1377612.1377620.

DAD is designed to facilitate the analysis and the comparisons of social welfare, inequality, poverty and equity across distributions of living standards and using disaggregated data. It is made available at no charge. DAD's features include the estimation of a large number of indices and curves that are useful for distributive comparisons as well as the provision of various statistical tools. It is currently the only software that systematically takes into account the sampling design of commonly used surveys in calculating asymptotic, bootstrap and p-bootstrap statistics for carrying out statistical inference. Many of DAD's features are useful for estimating the impact of programs (and reforms to these programs) on poverty and equity. DOI 10.3233/JEM-2009-0315

In order to investigate the output characteristic of a proton exchange membrane fuel cell (PEMFC) based on the electrical empirical model, a novel dynamic model of the PEMFC has been developed with MATLAB/Simulink, which is distinct from the models that have been published previously. By using a fuel cell test system of the Fuel Cell Application Centre (FAC) at Temasek Polytechnic, the transient electrical responses of PEMFC were conducted and analyzed under various operating conditions. A good match is found between simulation results and experimental data. The comprehensive results of simulation manifested that the model is effective and operational. This model will be very useful to optimize the structure design, improve the operation performance, and develop the real-time control system of PEMFC. Index Terms--Dynamic model, dynamic simulation, proton exchange membrane fuel cell (PEMFC), transient electrical response.

This article reviews three software packages that can be used to perform latent class cluster analysis, namely, Latent Gold[R], MCLUST, and poLCA. Latent Gold[R] is a product of Statistical Innovations whereas MCLUST and poLCA are packages written in R and are available through the web site http: //www.r-project.org. We use a single dataset and apply each software package to develop a latent class cluster analysis for the data. This allows us to compare the features and the resulting clusters from each software package. Each software package has its strengths and weaknesses and we compare the software from the perspectives of usability, cost, data characteristics, and performance. Whereas each software package utilizes the same methodology, we show that each results in a different cluster solution and suggest some rationales for deciding which package to use. KEY WORDS: Latent class models; Latent Gold[R]; MCLUST, Mixture models; poLCA.

All elections are historic, but this one was a doozy! The nomination and election of Barack Obama as president was dramatic, exciting and filled with surprises. Prior to the Iowa […]

We present here the algorithms and user interface of a MATLAB program, Fie, that solves numerically Fredholm integral equations of the second kind on an interval [a, b] to a specified, modest accuracy. The kernel function K(s, t) is moderately smooth on [a, b ] x [a, b] except possibly across the diagonal s = t. If the interval is finite, Fie provides for kernel functions that behave in a variety of ways across the diagonal, that is, K(s, t) may be smooth, have a discontinuity in a low-order derivative, have a logarithmic singularity, or have an algebraic singularity. Fie also solves a large class of integral equations with moderately smooth kernel function on [0, [infinity]). Categories and Subject Descriptors: G. 1.9 [Numerical Analysis]: Integral Equations; G.4 [Mathematical Software]: General Terms: Algorithms Additional Key Words and Phrases: Numerical solution, MATLAB

The speed of ultrasonic motor of piezo-electric type is usually measured using mechanical sensors such as pulse encoders. However, these sensors are costly and bulky. In this paper, a numerical speed estimation approach of a piezo-electric motor (PEM) is implemented using multilayer perception neural network (MLP-NN). The proposed model evaluates rotational speed and load torque based on the amplitude and driving frequency of the terminal voltage, considering the temperature variation. The estimated speed is employed to enhance the performance of the adaptive-fuzzy based speed control system. The model is validated and examined to achieve a minimized relative error in speed estimation approaches. Key words: Neural Network, MATLAB, USR60, hybrid model, INTRODUCTION Since the last two decades, speed-sensorless control methods of different motors using the estimated speed instead of measured speed have been studied. Recently, the use of neural networks NNs [...]

Teaching Mathematics to intermediate school boys is a challenging process which can greatly benefit from software technology and web-based learning. In this research we study the difficulties and opportunities of using this technology in teaching Mathematics to 13 year old boys studying Mathematics at school. For students' we identify the basic requirements at home, e.g., Mathematics computer software packages, availability of internet access and ability of the parents to provide computers, software and the internet for their sons. We assess the expected improvement in grades, motivation and communication between home and school. For the teaches', we assess the expected improvement in teachers' productivity, teachers' experience, qualifications in computing and Mathematics packages and their anticipated expectations such as improvement in problem solving, grades of students. Our main objective is to determine the expected results resulting from applying computer software and web-based learning in teaching Mathematics. Key words: Mathematics computer education, evaluation, learning and teaching, software, education, INTRODUCTION The introduction of information technology in recent years has provided the opportunity to integrate alternative learning techniques into traditional teaching (35). Learning or teaching systems have been developed in [...]

In this article we discuss a new software package, BACOLR, for the numerical solution of a general class of time-dependent 1-D PDEs. This package employs high-order adaptive methods in time and space within a method-of-lines approach and provides tolerance control of the spatial and temporal errors. The DAEs resulting from the spatial discretization (based on B-spline collocation) are handled by a substantially modified version of the Runge-Kutta solver, RADAU5. For each time step, the RADAU5 code computes an estimate of the temporal error and requires it to satisfy the user tolerance. After each time step BACOLR then computes a high-order estimate of the spatial error and requires this error estimate to satisfy the user tolerance. BACOLR was developed through a substantial modification of the adaptive method-of-lines package, BACOL. In this article we introduce the BACOLR package and present numerical results to show that the performance of BACOLR is comparable to and in some cases significantly superior to that of BACOL, which was shown in previous work to be more efficient, reliable and robust than other existing codes, especially for problems with solutions exhibiting narrow spikes or boundary layers. Categories and Subject Descriptors: G.1.0 [Numerical Analysis]: General Numerical algorithms; G.1.8 [Numerical Analysis]: Partial Differential Equations--Collocation methods, Method of lines, Runge-Kutta methods; G.4 [Mathematics of Computing]: Mathematics Software--Certification and testing, Efficiency General Terms: Algorithms, Performance Additional Key Words and Phrases: Numerical software, adaptive method-of-lines, 1-D PDEs, Runge-Kutta methods, differential-algebraic equations, spatial error control ACM Reference Format: Wang, R., Keast, P., and Muir, P. H. 2008. Algorithm 874: BACOLR: Spatial and temporal error control software for PDEs based on high-order adaptive collocation. ACM Trans. Math. Softw. 34, 3, Article 15 (May 2008), 28 pages. DOI = 10.1145/1356052.1356056 http://doi.acm.org/10.1145/1356052.1356056

Mass spectrometry based metabolomics represents a new area for bioinformatics technology development. While the computational tools currently available such as XCMS statistically assess and rank LC--MS features, they do not provide information about their structural identity. XCM[S.sup.2] is an open source software package which has been developed to automatically search tandem mass spectrometry (MS/MS) data against high quality experimental MS/MS data from known metabolites contained in a reference library (METLIN). Scoring of hits is based on a 'shared peak count' method that identifies masses of fragment ions shared between the analytical and reference MS/MS spectra. Another functional component of XCM[S.sup.2] is the capability of providing structural information for unknown metabolites, which are not in the METLIN database. This 'similarity search' algorithm has been developed to detect possible structural motifs in the unknown metabolite which may produce characteristic fragment ions and neutral losses to related reference compounds contained in METLIN, even if the precursor masses are not the same.

Direct torque control (DTC) is one of the most excellent control strategies of torque control in induction machine. It is considered as an alternative to the field oriented control (FOC) or vector control technique. These two control strategies are different on the operation principle but their objectives are the same. They aim to control effectively the torque and flux. Torque control of an induction machine based on DTC strategy has been developed and a comprehensive study is present in this research. The performance of this control method has been demonstrated by simulations performed using a versatile simulation package, Matlab/Simulink. Several numerical simulations have been carried out in a steady state and transient operation on a speed control mode. Key words: Direct torque control, induction machine, vector control, Matlab/Simulink, INTRODUCTION For many years, induction machine have provided the most common form of electromechanical drive for industrial, commercial and domestic applications that can operate at essentially constant speed. Induction machines [...]

A MATLAB 6.0 implementation of the LSTRS method is presented. LSTRS was described in Rojas et al. [2000]. LSTRS is designed for large-scale quadratic problems with one norm constraint. The method is based on a reformulation of the trust-region subproblem as a parameterized eigenvalue problem, and consists of an iterative procedure that finds the optimal value for the parameter. The adjustment of the parameter requires the solution of a large-scale eigenvalue problem at each step. LSTRS relies on matrix-vector products only and has low and fixed storage requirements, features that make it suitable for large-scale computations. In the MATLAB implementation, the Hessian matrix of the quadratic objective function can be specified either explicitly, or in the form of a matrix-vector multiplication routine. Therefore, the implementation preserves the matrix-free nature of the method. A description of the LSTRS method and of the MATLAB software, version 1.2, is presented. Comparisons with other techniques and applications of the method are also included. A guide for using the software and examples are provided. Categories and Subject Descriptors: G.4 [Mathematical Software]--Algorithm design and analysis General Terms: Algorithms, Design

We became aware of a few issues related to our recent review of statistical software packages that are devoted primarily to or contain a strong component of exact statistical tests and procedures (Oster and Hilbe 2008) after our review had already gone to press. These issues are focused on our evaluation of a few specific components of SAS and SPSS that involve exact and asymptotic statistical analyses. We address each of these issues in this article. KEY WORDS: Categorical data analysis; Statistical computing; Statistical software testing.

The use of symbolic calculators such as MATHEMATICA is becoming more commonplace among upper level physics students. The presence of such powerful calculators can couple strongly to the type of mathematical reasoning students employ. These tools do not merely offer students a convenient way to perform the calculations they would have otherwise done by hand. We present examples from the work of upper level physics majors where MATHEMATICA plays an active role in focusing and sustaining their thoughts around calculation. These students still engage in powerful mathematical reasoning while they calculate, but struggle because of the narrowed breadth of their thinking. We model MATHEMATICA'S influence as an integral part of the constant feedback that occurs in how students frame, and hence focus, their work. [DOI: 10.1119/1.2835053]

This review evaluates five statistical software packages that are devoted primarily to, or contain a strong component in, exact statistical tests and procedures. The five packages are LogXact, SAS, SPSS, Stata, and StatXact. Only Windows versions of these packages are evaluated. They are examined according to various criteria that should be of interest to anyone who needs to perform either exact or asymptotic parametric or nonparametric data analyses. It is shown that the software packages differ in their capabilities for exact methods, in how these methods are implemented, and in their potential user audiences. Recommendations for users and developers of these software packages are provided. KEY WORDS: Statistical computing; Categorical data analysis; Statistical software testing

IN THE DECEMBER 2003 ISSUE of OR/MS Today, I reviewed Mathematica 5 from an operations research professional's viewpoint [1]. I wrote how version 5 is compelling for O.R. professionals who […]

A number of studies have investigated the numerical reliability of the GAUSS Mathematical and Statistical System by Aptech Inc. Those studies identified several critical accuracy errors for different computational methods. We conducted comprehensive tests of this widely used package on estimation, statistical distributions, and random number generation and found that GAUSS 8.0 still has serious problems, especially with statistical distributions and random number generation. KEY WORDS: Software testing; Numerical accuracy; Statistical reference datasets (StRD).

Power quality is an important measure of the performance of an electrical power system. This paper discusses the topology, control strategies using artificial intelligent based controllers and the performance of a unified power quality conditioner for power quality improvement. UPQC is an integration of shunt and series compensation to limit the harmonic contamination within 5 %, the limit imposed by IEEE-519 standard. The novelty of this paper lies in the application of neural network control algorithms such as model reference control and Nonlinear Autoregressive-Moving Average (NARMA)-L2 control to generate switching signals for the series compensator of the UPQC system. The entire system has been modeled using MATLAB 7.0 toolbox. Simulation results demonstrate the applicability of MRC and NARMA-L2 controllers for the control of UPQC. Key words: APF, MRC, NARMA-L2, VSI, DVR, INTRODUCTION The better controllability, higher efficiency, higher current carrying capability, and fast switching characteristics of static power converters are promoting major changes in controlling the power flow of transmission and [...]

Abstract: In this paper, we discuss a Mathematica program that we have written that calculates the integrand for amplitudes in the closed-time-path formulation of real-time finite-temperature field theory. The program [...]

The modern world is full of uncertainty and variation. Last summer the cost of gasoline varied almost daily, subject to the interplay of demand, current stock levels and interdictions of […]

WHEN THE SOFTWARE REVIEW EDITOR for OR/MS Today is looking for someone to review software, I usually lock my office door and hide under the desk until I get the […]

In the 1990's, Dhillon and Parlett devised the algorithm of multiple relatively robust representations (MRRR) for computing numerically orthogonal eigenvectors of a symmetric tridiagonal matrix T with O([n.sup.2]) cost. While previous publications related to MRRR focused on theoretical aspects of the algorithm, a documentation of software issues has been missing. In this article, we discuss the design and implementation of the new MRRR version STEGR that will be included in the next LAPACK release. By giving an algorithmic description of MRRR and identifying governing parameters, we hope to make STEGR more easily accessible and suitable for future performance tuning. Furthermore, this should help users understand design choices and tradeoffs when using the code. Categories and Subject Descriptors: G.1.3 [Numerical Analysis]: Numerical Linear Algebra General Terms: Algorithms, Design Additional Key Words and Phrases: Multiple relatively robust representations, eigenvalues, eigenvectors, symmetric matrix, LAPACK, numerical software, design, implementation

A translation to Fortran 90 of Gertrude Blanch's algorithm for computing the expansion coefficients of the series that represent Mathieu functions is presented. Its advantages are portability, higher precision, practicality of use, and extended documentation. In addition, numerical validations and comparisons with other existing methods are presented. Categories and Subject Descriptors: G.4 [Mathematics of Computing]: Mathematical Software--Documentation; J.2 [Computer Applications]: Physical Sciences and Engineering--Engineering, mathematics and statistics, physics General Terms: Algorithms, Documentation, Languages, Verification Additional Key Words and Phrases: Mathieu function, special function, computation, validation

BABDCR is a package of Fortran 90 subroutines for the solution of linear systems with bordered almost block diagonal coefficient matrices. It is designed to handle matrices with blocks of the same size, that is, having a block upper bidiagonal structure with an additional block in the right upper corner. The algorithm implemented in the package performs cyclic reduction of the coefficient matrix in order to reduce the fill-in due to the corner block. Categories and Subject Descriptors: G.1.3 [Numerical Analysis]: Numerical Linear Algebra--Sparse, structured, and very large systems (direct and iterative methods); G.4 [Mathematics of Computing]: Mathematical Software--Algorithm design and analysis; documentation General Terms: Algorithms, Documentation Additional Key Words and Phrases: Linear systems, bordered almost block diagonal matrices, numerical solution, cyclic reduction

We present an algorithm to compute integrals of the form [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] with [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (x) the Bessel function of the first kind and (real) order [v.sub.i]. The parameter m is a real number such that [[summation].sub.i] [v.sub.i] + m > -1 and the coefficients [a.sub.i] are strictly positive real numbers. The main ingredients in this algorithm are the well-known asymptotic expansion for [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.](x) and the observation that the infinite part of the integral can be approximated using the incomplete Gamma function [GAMMA](a, z). Accurate error estimates are included in the algorithm, which is implemented as a MATLAB program. Categories and Subject Descriptors: G. 1.4 [Numerical Analysis]: Quadrature and Numerical Differentiation; G.4 [Mathematics of Computing]: Mathematical Software General Terms: Algorithms Additional Key Words and Phrases: Quadrature, Bessel functions

Globally convergent, probability-one homotopy methods have proven to be very effective for finding all the isolated solutions to polynomial systems of equations. After many years of development, homotopy path trackers based on probability-one homotopy methods are reliable and fast. Now, theoretical advances reducing the number of homotopy paths that must be tracked and handling singular solutions have made probability-one homotopy methods even more practical. POLSYS_GLP consists of Fortran 95 modules for finding all isolated solutions of a complex coefficient polynomial system of equations. The package is intended to be used on a distributed memory multiprocessor in conjunction with HOMPACK90 (Algorithm 777), and makes extensive use of Fortran 95-derived data types and MPI to support a general linear product (GLP) polynomial system structure. GLP structure is intermediate between the partitioned linear product structure used by POLSYS_PLP (Algorithm 801) and the BKK-based structure used by PHCPACK. The code requires a GLP structure as input, and although finding the optimal GLP structure is a difficult combinatorial problem, generally physical or engineering intuition about a problem yields a very good GLP structure. POLSYS_GLP employs a sophisticated power series end game for handling singular solutions, and provides support for problem definition both at a high level and via hand-crafted code. Different GLP structures and their corresponding Bezout numbers can be systematically explored before committing to root finding. Categories and Subject Descriptors: G.1.5 [Numerical Analysis]: Roots of Nonlinear Equations--Continuation (homotopy) methods, polynomials, methods for systems of equations; G.4 [Mathematics of Computing]: Mathematical Software General Terms: Algorithms Additional Key Words and Phrases: Chow-Yorke algorithm, curve tracking, fixed point, Fortran 95, general linear product, globally convergent, homotopy methods, linear product decomposition, probability-one, zero

APPSPACK is software for solving unconstrained and bound-constrained optimization problems. It implements an asynchronous parallel pattern search method that has been specifically designed for problems characterized by expensive function evaluations. Using APPSPACK to solve optimization problems has several advantages: No derivative information is needed; the procedure for evaluating the objective function can be executed via a separate program or script; the code can be run serially or in parallel, regardless of whether the function evaluation itself is parallel; and the software is freely available. We describe the underlying algorithm, data structures, and features of APPSPACK version 4.0, as well as how to use and customize the software. Categories and Subject Descriptors: J.2 [Computer Applications]: Physical Sciences and Engineering--Mathematics and statistics General Terms: Algorithms, Design, Documentation Additional Key Words and Phrases: Parallel derivative-free optimization, pattern search

We present a new out-of-core sparse symmetric-indefinite factorization algorithm. The most significant innovation of the new algorithm is a dynamic partitioning method for the sparse factor. This partitioning method results in very low I/O traffic and allows the algorithm to run at high computational rates, even though the factor is stored on a slow disk. Our implementation of the new code compares well with both high-performance in-core sparse symmetric-indefinite codes and a high-performance out-of-core sparse Cholesky code. Categories and Subject Descriptors: G.1.3 [Numerical Analysis]: Numerical Linear Algebra--Linear systems (direct and iterative methods); sparse, structured, and very large systems (direct and iterative methods); G.4 [Mathematical Software]: Algorithm design and analysis; G.4 [Mathematics of Computing]: Mathematical Software--Efficiency; E.5 [Data]: Files--Organization/structure General Terms: Algorithms, Performance, Experimentation Additional Key Words and Phrases: Out-of-Core, symmetric-indefinite

This article offers a review of three software packages that estimate directed acyclic graphs (DAGs) from data. The three packages, MIM, Tetrad and WinMine, can help researchers discover underlying causal structure. Although each package uses a different algorithm, the results are to some extent similar. All three packages are free and easy to use. They are likely to be of interest to researchers who do not have strong theory regarding the causal structure in their data. DAG modeling is a powerful analytic tool to consider in conjunction with, or in place of, path analysis, structural equation modeling, and other statistical techniques. KEY WORDS: Bayesian networks; Causality; Data mining; Indirect effects.