When I use the plsda function in mixOmics, the loading values are different from the ones I got from pls package. But when I compare with the loading values in plsda, it looks like the values are more similar to the loading.weight in the pls package. Could you help me to figure out why? Are the loading values in mixOmics same as the loading.weight in pls package? And why go with this way?
Thanks a lot!
We have an additional outputs called
loadings.star (which has not been mentioned in the help file, we will fix this).
loadings.star are the loadings recalculated based on the original matrix, whereas
loadings are the loadings calculated based on the deflated (residual) matrix across the dimension of the PLS model.
It could be that the outputs from the package pls only output the loading.weigth = loading.star, you can investigate or read their help files.