Hello,
I’ve a question about the vip() function. I’ve my model already optimized and then I calculate the vips through vip() function.
My question is whether each vip score accumulates in the components or are they separated?
In order to have a single VIP value for each variable, across all components, how can I do this?
Is there a way to do it with this function or do I have to apply a formula to aggregate them?
hi @Sabribar
You can have a look at this: VIP by groups in PLS-DA - #6 by kimanh.lecao for the formula of the VIP. Each VIP score accumulates across components already (i.e if you consider 4 components, then the value of the VIP on the 4th component includes the information from the previous components).
Kim-Anh
Thank you!! This is helpfull for my analysis