Entering edit mode
Hi all, I run a permutation test on opls-da and my Q2Y was significant. This was done with 300 permutations. I need to actually obtain the 300 values of Q2Y that were compared to the value of Q2Y obtained from non-permuted model. The code below automatically gives p-value and diagnostic plots but I"m also intrested in the 300 values of Q2Y. I would appreciate it if you could help with the code for that. Thank you
perm_Y_CE<-Y_CE[c(1:100)]
CEperm_oplsda <- opls(CEperm, perm_Y_CE, scaleC="standard",predI = 1, orthoI = 3,permI=300)
The
oplsobject contains the permMN matrix in the supplementary list (suppLsslot):The first row of permMN corresponds to the model without response permutation. The following rows correspond to the permutations. The sim column contains the values of the similarity between the permutated and the original response, and the Q2(cum) column gives the Q2 values.