How to get feature importance in xgboost?

Question

I\'m using xgboost to build a model, and try to find the importance of each feature using get_fscore(), but it returns {}

and my train code

Answer 1

Try this

fscore = clf.best_estimator_.booster().get_fscore()