[R-sig-eco] MARS model predicting null

[Rajendra Mohan Panda]

Dear All

My pleasure to be in this elite group of R users.

This time my query relates to multivariate regression spline (MARS). While
using MARS for a set of 37, 63 and 100 number of samples (in rows) with 94,
147 and 265 species (in columns) respectively, and with a list of 24
predictors, I get NULL value with RSQ zero for these models.

Kindly help me understand why this happens even if the predictors selected
are ecologically significant and selected after removing collinearity (by
'usdm').

Best Regards
Rajendra M Panda
School of Water Resources
Indian Institute of Technology Kharagpur

	[[alternative HTML version deleted]]