Issue in using predict() with a plsda model

[Former user]

Hello,

I have an error after using predict() on a plsda model. The creation of the model and using plotIndiv and plotVar is OK but when I try to use my model to predict, I get this error :

Error in solve.default(Sr) : system is computationally singular: reciprocal condition number = 1.32942e-18

Could anyone tell me why I get this error and how can I solve it ?

Thks

[Florian Rohart]

Hi there,

There is an inverse calculation in the predict function, that seems to be what's causing this problem for you. Unfortunately, this is very data dependent.

How many components in your plsda model? Maybe there is too many components and inversing the matrices becomes tricky..