Dears,
Here is Alessio.
I am currently trying to include PLS analysis from MixOmics in my workflow.
Usually, I plot R2 and Q2 values to determine number of component to select and assess the quality of the model, as described in a textbook from UMETRICS ACADEMY.
However, Q2.total from mixOmics is often higher than R2, whereas I never found it in the examples from UMETRICS. Moreover I performed the same analysis in SIMCA and both R2 and Q2 are nicely distributed. In number, mixOmics 1st component have Q2.total = 0.85 whereas, SIMCA gives Q2cum = 0.14.
Are Q2.total and Q2cum a different values?