Supporting bootstrap sampling as in sklearn

[hmishfaq]

I am trying to convert a sklearn based code that uses sklearn.ensemble.BaggingRegressor and am wondering if Ensemble-pytorch can have bootstrap argument as sklearn has.

[xuyxu]

Hi @hmishfaq, could you further explain which argument do you want to use. Currently, the Bagging estimators in Ensemble-pytorch only use the sample data with replacement technique.

[hmishfaq]

Hi @xuyxu, I would like to use the argument bootstrap as described in the link above of sklearn.ensemble.BagginRegressor. However, it seems there is no analog of such argument in ensemble-pytorch.

bootstrap:bool, default=True
Whether samples are drawn with replacement. If False, sampling without replacement is performed.

[xuyxu]

This argument equals True in BaggingRegressor internally.