I was getting this error when trying to run gbm.fit:
The data set is too small or the subsampling rate is too large: nTrain * bag.fraction <= n.minobsinnode.
I modified my arguments to ensure nTrain * bag.fraction > n.minobsinnode. However, I was still getting the error. Upon closer examination of the source code, I see here that the error is generated when nTrain * bag.fraction <= 2 * n.minobsinnode + 1. This error is different that the one reported in the message, so I suggest correcting the inequality in the error message.
gregridgeway
commented
8 months ago
Hello,
I was getting this error when trying to run
gbm.fit:The data set is too small or the subsampling rate is too large: nTrain * bag.fraction <= n.minobsinnode.I modified my arguments to ensure
nTrain * bag.fraction > n.minobsinnode. However, I was still getting the error. Upon closer examination of the source code, I see here that the error is generated whennTrain * bag.fraction <= 2 * n.minobsinnode + 1. This error is different that the one reported in the message, so I suggest correcting the inequality in the error message.