Linear regression based on linear structure between covariates

[Up] [Top]

Documentation for package ‘CorReg’ version 1.0

Help Pages

CorReg-package Algorithms for regression with correlated covariates
BicZ Compute the BIC of a given structure
BicZcurve Curve of the BIC for each possible p2 with a fixed Z and truncature of Z
cleancolZ clean Z columns (if BIC improved)
cleanYtest Selection method based on p-values (coefficients)
cleanZ clean Z (if BIC improved)
cleanZR2 To clean Z based on R2
compare_beta compare signs of the coefficients in two vectors
compare_sign compare signs of the coefficients in two vectors
compare_struct To compare structures (Z)
compare_zero compare 0 values in two vectors
confint_coef plot and give confidence intervals on the coefficients estimated in a model or for proportions
correg Estimates the response variable using a structure
CVMSE Cross validation
density_estimation BIC of estimated marginal gaussian mixture densities
Estep Imputation of missing values knowing alpha (E step of the EM)
fillmiss Fill the missing values in the dataset
hatB Estimates B matrix
matplot_zone draws matplot with conditionnal background for easier comparison of curves.
mixture_generator Gaussiam mixture dataset generator with regression between the covariates
MSEZ Computes the MSE on the joint distribution of the dataset
MSE_loc simple MSE function
OLS Ordinary Least Square efficiently computed with SEM for missing values
ProbaZ Probability of Z without knowing the dataset. It also gives the exact number of binary nilpotent matrices of size p.
R2Z Estimates R2 of each subregression
readY a summary-like function
readZ read the structure and explain it
recursive_tree decision tree in a recursive way
rforge Upgrades a package to the lastest version on R-forge
searchZ_sparse Sparse structure research
showdata show the missing values of a dataset
structureFinder MCMC algo to find a structure between the covariates
Terminator Destructing values to have missing ones
WhoIs Give the partition implied by a structure
Winitial initialization based on a wheight matrix (correlation or other)
Y_generator Response variable generator with a linear model