How to do Multivariate Adaptive Regression Splines feature selection in python? Specifically, I need the python equivalent of the earth function in R

Question

This is the code in R:

marsModel <- earth(eval(parse(text=paste(ResponseVariable,"~."))), data = data) #build model
ev <- evimp (marsModel)

Response variable is self-explanatory, the data contains 178 features and 1 response variable. Each feature has 56 observations.

'ev' ends up with 10 features selected out of 178 and their gcv (Generalized cross validation), rss (Residual sum of squares) and nsubsets (no. of subset models the variable occurs) scores.