Strange values in prediction and ensemble with Mahalanobis and glmnet methods

[mauriciovancine]

Dear Babak,

How are you?

First I would like to congratulate you for the sdm package, certainly a huge step forward in standardizing the use of SDMs.

I would like to ask about some strange values I am getting using the Mahalanobis and glmnet algorithms.

When I plot the predicted from other methods all fall between 0 and 1. For these two, there are negative values.

When I build the ensemble with these methods, the final values are also negative.

Thank you very much.

Mauricio Vancine