Cross validation of OPLS-DA and PLS regression models

Question

I obtained for the first time strange results from my OPLS-DA and PLS regression models, the CV ANOVA was (-nan(ind)), the SD residual was (inf), and the MS residual was (inf), moreover, the F test was equal to zero.

I don't know if I have a problem with my results?

Answer 1

One thing I would do is check your data. I've seen results like this when there are data cleaning issues such as variables with low/no variance or large proportions of missing data in individual variables.