Warnings for GLM/RF/BRT/MARS /Maxent

[virginia...@gmail.com]

Hello!

I've been consistently getting the following warnings when running different species occurrence datasets through the 5 different statistical modules (GLM/RF/BRT/MARS/Maxent). I was just wondering if these are warnings that I should be worried about in terms of interpreting the model output, tuning the module settings, or there being problems with my occurrence/environmental data:

1) glm.fit: algorithm did not converge

2) glm.fit: fitted probabilities numerically 0 or 1 occurred (this warning is particularly common for MARS)

3) In cor(auc.data$pres.abs, x$pred) : the standard deviation is zero

Thank you!!

Virginia