Chapter 2: Regression

Routines

Simple Linear Regression

Straight line fit............................................................................................................ RLINE

Simple linear regression analysis................................................................................. RONE

Response control by a fitted line.................................................................................. RINCF

Inverse prediction by a fitted line.................................................................................. RINPF

Multivariate General Linear Model Analysis

Model Fitting

From raw data for a single dependent variable................................................................ RLSE

From covariances....................................................................................................... RCOV

From raw data without classification variables.............................................................. RGIVN

From raw data with classification variables.................................................................... RGLM

With linear equality restrictions.................................................................................. RLEQU

Statistical Inference and Diagnostics

Summary statistics for a fitted regression.................................................................... RSTAT

Variance-covariancematrix of the estimated coefficients............................................... RCOVB

Construction of a completely testable hypothesis.......................................................... CESTI

Sums of crossproducts for a multivariate hypothesis.................................................... RHPSS

Tests for the multivariate linear hypothesis.................................................................. RHPTE

Test for lack of fit based on exact replicates................................................................ RLOFE

Test for lack of fit based on near replicates.................................................................. RLOFN

Intervals and diagnostics for individual cases............................................................... RCASE

Diagnostics for outliers and influential cases................................................................. ROTIN

Utilities for Classification Variables

Getting unique values of classification variables........................................................... GCLAS

Generation of regressors for a general linear model..................................................... GRGLM

Variable Selection

All best regressions via leaps-and-bounds algorithm..................................................... RBEST

Stepwise regression.................................................................................................. RSTEP

Generalized sweep of a nonnegative definite matrix.................................................... GSWEP

Retrieval of a symmetric submatrix from a symmetric matrix........................................ RSUBM

Polynomial Regression and Second-Order Models

Polynomial Regression Analysis

Polynomial fit of known degree.................................................................................. RCURV

Polynomial regression analysis.................................................................................. RPOLY

Second-Order Model Design

Generation of an orthogonal central composite design................................................. RCOMP

Utility Routines for Polynomial Models and Second-Order Models

Polynomial regression fit........................................................................................... RFORP

Summary statistics for a fitted polynomial model......................................................... RSTAP

Case statistics for a fitted polynomial model............................................................... RCASP

Generation of orthogonal polynomials......................................................................... OPOLY

Centering of variables and generation of crossproducts................................................ GCSCP

Transforming coefficients for a second order model...................................................... TCSCP

Nonlinear Regression Analysis

Nonlinear regression fit............................................................................................... RNLIN

Fitting Linear Models Based on Criteria Other Than Least Squares

Least absolute value regression.................................................................................... RLAV

Least Lp norm regression............................................................................................. RLLP

Least maximum value regression................................................................................. RLMV

Partial Least Squares Regression................................................................................. PLSR



http://www.vni.com/
PHONE: 713.784.3131
FAX:713.781.9260