Learn R Programming

armada (version 0.1.0)

A Statistical Methodology to Select Covariates in High-Dimensional Data under Dependence

Description

Two steps variable selection procedure in a context of high-dimensional dependent data but few observations. First step is dedicated to eliminate dependence between variables (clustering of variables, followed by factor analysis inside each cluster). Second step is a variable selection using by aggregation of adapted methods. Bastien B., Chakir H., Gegout-Petit A., Muller-Gueudin A., Shi Y. A statistical methodology to select covariates in high-dimensional data under dependence. Application to the classification of genetic profiles associated with outcome of a non-small-cell lung cancer treatment. 2018. .

Copy Link

Version

Install

install.packages('armada')

Monthly Downloads

26

Version

0.1.0

License

GPL-3

Maintainer

Aurelie Gueudin

Last Published

April 4th, 2019

Functions in armada (0.1.0)

covariables

concatenation of the rownames of X and of the response vector Y.
X_decor

Decorrelation of a matrix X, given a response variable Y.
clustering

To obtain the dendrogram of the covariates contained in the data.frame X, and a proposition for the number of clusters of covariates in X.
ARMADA.select

Covariates selection via 8 selection methods
toys.data.reg

Toys data in regression case
ARMADA.summary

Scores of the covariates X
ARMADA

Scores of all the covariates present in X, given the vector Y of the response.
ARMADA.heatmap

Heatmap of the selected covariates.
toys.data.multi

Toys data in multinomial case
toys.data

Toys data