Problem with Cubist Predict function. Predicting 0 for large data.

Open kchaitanyabandi opened this issue 8 years ago • 1 comments

Hi,

The cubist prediction isn't working properly for large datasets (about 10,000 rows). It works completely fine if I predict row by row, but gives all zeroes when I pass all the 10,000 rows or any subset of the data (< 10,000 rows). I think this is a memory issue. Could this be fixed? We have a really important implementation on hold because of this issue.

Thanks Krishna Chaitanya Bandi

Dec 21 '17 06:12 kchaitanyabandi

Can you give a reproducible example (even if it is artificial or blinded) and the results of sessionInfo()?

May 17 '18 19:05 topepo