When one of the columns of the target matrix Y is constant the following error is raised both for plsca and plsreg2 :

Error in if (ah.dif < 1e-06 || iter == 100) break : 
  missing value where TRUE/FALSE needed

The following code snippet will allow to reproduce this:

data(linnerud)
X = linnerud[4:6]
Y = linnerud[1:3]
Y[,2] = 1
plsreg2(X, Y)
plsca(X, Y)

Is this behaviour expected?

Hi, nice package. However, there is some problem to build the package from source:
C:\Program Files\R\R-3.2.0\bin\x64>rcmd INSTALL "C:/R/Package clones/plsdepot"

• installing to library 'c:/R/win-library/3.2'
• installing source package 'plsdepot' ...
  ** R
  Error in .install_package_code_files(".", instdir) :
  files in 'C:/R/Package clones/plsdepot/R' missing from 'Collate' field:
  plsdepot.R
  ERROR: unable to collate and parse R files for package 'plsdepot'
• removing 'c:/R/win-library/3.2/plsdepot'

Is there any way I can calculate VIP scores using plsreg1?

Scaling should applied separately on the training set and the test set (using the mean and sd from the training set).
The current code provides an over-optimistic result.

Dear author of plsdepot package,
Sometimes, its is interesting to explore how individual observations relate to the X and Y variables analyzed.
In this regard that would be interesting to project the observations on the correlation plot, as suggested in this thread (https://stackoverflow.com/questions/52906389/correlation-loading-plot-from-plsr-with-observations-using-ggplot2). The plot example in the thread was produced from a commercial software. I tried to reproduce a similar plot using pls/plsdepot packages and ggplot.
I guess that would require a transformation step of both the x.scores and y.scores so that individual observations (rows) can be projected on a correlation loading scale (-1: 1). I haven't yet found a solution to that. Any suggestions will be much appreciated.

Hello,

I have used the plsreg2() function to create a model to predict rangeland condition using various remote sensing variables. I have, however, been unable to determine a way to predict the model over my project area. It seems that the predict() function does not exist within the package. Any suggestions on how to predict the model would be highly appreciated.

Thank you and apologies if I am misunderstanding the package use!