Learn R Programming

CorReg (version 1.2.17)

Linear Regression Based on Linear Structure Between Variables

Description

Linear regression based on a recursive structural equation model (explicit multiples correlations) found by a M.C.M.C.(Markov Chain Monte Carlo) algorithm. It permits to face highly correlated variables. Variable selection is included (by lasso, elastic net, etc.). It also provides some graphical tools for basic statistics. For more information about the method, read the PhD thesis of Clement Thery (2015) in the link below.

Copy Link

Version

Install

install.packages('CorReg')

Monthly Downloads

197

Version

1.2.17

License

CeCILL

Homepage

http://www.correg.org

Maintainer

Last Published

February 20th, 2020

Functions in CorReg (1.2.17)

Destructing values to have missing ones

Read the structure and explain it

Linear regression using CorReg's method, with variable selection.

Clean Z's columns based on p-values (coefficients or global)

clean the structure of correlations Z (if BIC improved)

To compare sub-regression structures

Replaces unwanted values by NAs

density_estimation

BIC of estimated marginal gaussian mixture densities

Matplot with curves comparison by background colors.

mixture_generator

Gaussian mixtures dataset generator with regression between the covariates

How would it be if we were naive ?

Decision tree in a recursive way

To show the missing values of a dataset

Quickly reports some MSE

structureFinder

MCMC algorithm to find a structure between the covariates

Removes missing values (rows and column to obtain a large full matrix)

To clean non numeric values in a vector

Compute the BIC of a given structure

Probability of Z without knowing the dataset. It also gives the exact number of binary nilpotent matrices of size p.

Quick tutorial for CorReg package

Boxplot with confidence interval and ANOVA on the plot.

Histograms with clusters

Simple MSE function

Cross validation