Any advantage of sklearn wrappers for xgboost over python API?

Question

Are there any advantages of using the XGBoost sklearn wrappers XGBRegressor or XGBClassifier over using the Python API with the train method? Is one faster than the other or does one use less memory than the other? I am asking for a generic case where the data comes in some form which I convert to a pandas dataframe.

Answer 1

One advantage of using scikit-learn wrappers is that they are more likely to work within the scikit-learn ecosystem. For example, it might be easier to put the algorithm into pipelines.