calculate_PLS_variances_explained: Gets the variances explained from predictor LVs through...

Description Usage Arguments Value

View source: R/calculate_PLS_variances_explained.R

Description

Variance explained is calculated by looking at the ratio between the Sum of Squares of the difference between Y and predicted Y through the direct effects, and the Sum of Squares of Y. This ratio is subtracted from 1 to yield the variance explained between 0 and 1. Note that negative values are possible if more variance is introduced by prediction. This indicates that a certain effect is not predictive, and one should consider to remove the connection in the model definition.

Usage

1
calculate_PLS_variances_explained(model, scaling = "numerical")

Arguments

model

A path_model model estimated using the process_pls wrapper.

Value

a matrix, the same shape as the connection matrix, where the non zero elements indicate the explained variance from one node to another. The matrix is lower triangular.


GeertPostma/pathmodelr documentation built on Oct. 5, 2021, 4:17 p.m.