GoreLab / waves

R package for Vis-NIR spectral analysis
https://gorelab.github.io/waves/
Other
6 stars 6 forks source link

PLS variable importance #9

Closed jmh579 closed 2 years ago

jmh579 commented 3 years ago

Implement partial least squares variable importance in addition to random forest variable importance

jmh579 commented 3 years ago

caret varImp documentation:

"Partial Least Squares: the variable importance measure here is based on weighted sums of the absolute regression coefficients. The weights are a function of the reduction of the sums of squares across the number of PLS components and are computed separately for each outcome. Therefore, the contribution of the coefficients are weighted proportionally to the reduction in the sums of squares."

Example: gbmImp <- varImp(gbmFit3, scale = FALSE) where gbmFit3 is a trained model

jmh579 commented 2 years ago

Addressed in 0.2.0