Learn R Programming

⚠️There's a newer version (1.6.2) of this package.Take me there.

CorrBin (version 1.5)

Nonparametrics with Clustered Binary and Multinomial Data

Description

This package implements non-parametric analyses for clustered binary and multinomial data. The elements of the cluster are assumed exchangeable, and identical joint distribution (also known as marginal compatibility, or reproducibility) is assumed for clusters of different sizes. A trend test based on stochastic ordering is implemented.

Copy Link

Version

Install

install.packages('CorrBin')

Monthly Downloads

207

Version

1.5

License

GPL (>= 2)

Maintainer

Aniko Szabo

Last Published

January 5th, 2015

Functions in CorrBin (1.5)

Internal CorrBin objects

Likelihood ratio test of stochastic ordering

Create a `CBdata' object from a data frame.

GEE-based trend test

Functions for generating multinomial outcomes

Test for increasing trend with correlated binary data

Estimate joint event probabilities for multinomial data

Extract from a CBData or CMData object

Control values for order-constrained fit

Developmental toxicology study of DEHP in mice

Test the assumption of marginal compatibility

Extract correlation coefficients from joint probability arrays

Read data from external file into a CMData object

CorrBin-package

Nonparameterics for Correlated Binary and Multinomial Data

Likelihood-ratio test statistic

Unwrap a clustered object

Finding the NOSTASOT dose

Read data from external file into a CBData object

Generate a random CMData object

Parametric distributions for correlated binary data

Extract univariate marginal probabilities from joint probability arrays

Generate random correlated binary data

Rao-Scott trend test

Distribution of the number of responses assuming marginal compatibility.

Create a `CMdata' object from a data frame.

Order-restricted MLE assuming marginal compatibility

Shell Toxicology data