Q2 for Y-variables in PLS canonical mode

Analysis
1

Hello

I am using the perf function applied on PLS (canonical mode). Results show Q2.total, but not the Q2 for each of the Y variables. Is it possible to get these values?

Thank you in advance

2

hi @Joe,

PLS canonical mode is a tricky one as we don’t perform a regression. That’s why you don’t have the Q2 for the Y variables. The fact that you are using the Q2.total at all make me think that maybe you are interested in a regression instead? (see 4 PLS on the liver toxicity study | mixOmics vignette)

Otherwise you can also consider rCCA, if you are not interested in variable selection.
Overall PLS canonical mode is primarily to check if the datasets are integrable / include common information but we have not proposed Manny performance measures to go beyond this observation.

Kim-Anh