Learn R Programming
⚠️There's a newer version (2.0.0) of this package.Take me there.

automultinomial (version 1.0.0)

Autologistic and Automultinomial Spatial Regression and Variable Selection

Description

Contains functions for autologistic variable selection and parameter estimation for spatially correlated categorical data (including k>2). The main function is MPLE. Capable of fitting the centered autologistic model described in Caragea and Kaiser (2009), as well as the traditional autologistic model of Besag (1974).

Copy Link

Version

Install

install.packages('automultinomial')

Monthly Downloads

28

Version

1.0.0

License

GPL-2 | GPL-3

Maintainer

Stephen Berg

Last Published

October 5th, 2016

Functions in automultinomial (1.0.0)

JMatrixCentered

J matrix for centered variance estimate
adjMat

Spatial adjacency matrix
centeredLogLikHess

Centered model hessian
logLikelihood

Compute log likelihood
criterionFit

Minimize criterion (AIC or BIC)
significanceTest

Significance tests for autologistic/automultinomial models
sandwichVariance

Compute sandwich variance estimate
plotGrid

Visualize response values on a grid
totalX

Add columns to X matrix
MPLE

Logistic autoregression
multinomialBCD

Block coordinate descent
multProb

Multinomial probabilities
MPLEinternal

Multinomial regression
simulateData

Simulate data
uncenteredLogLikGrad

Uncentered gradient
unStandardize

Unstandardize coefficients
uncenteredLogLikHess

Uncentered hessian
chol_psd

Cholesky decomposition for positive semidefinite matrix
centeredLogLikelihood

Centered model log pseudolikelihood
indsFunction

Get indices of relevant component of theta matrix
JMatrixUncentered

Uncentered J matrix
centeredLogLikGrad

Centered model gradient