regcoeffs.getStats: Distribution statistics for regression coeffificents

In mdatools: Multivariate Data Analysis for Chemometrics

Distribution statistics for regression coeffificents

Description

calculates standard error, t-values and p-values for regression coefficients based on Jack-Knifing method.

Usage

regcoeffs.getStats(coeffs, ci.coeffs = NULL, use.mean = TRUE)

Arguments

coeffs	array (npred x ncomp x nresp) with regression coefficients
ci.coeffs	array (npred x ncomp x nresp x cv) with regression coefficients for computing confidence intervals (e.g. from cross-validation) using Jack-Knifing method
use.mean	logical, tells how to compute standard error for regression coefficients. If TRUE mean values for ci.coeffs is computed first. If FALSE, values (coefficients computed for global model) are used as mean.

Value

a list with statistics three arrays: srandard error, t-values and p-values computed for each regression coefficient.

mdatools documentation built on Sept. 11, 2024, 7:59 p.m.