###xgboost在做Feature Importance图时候，新旧版存在差异https://blog.csdn.net/u013313168/article/details/80911422旧版用alg.booster().get_fscore()新版用alg.get_booster().get_fscore()否则会报错  File "
    
     ", line 1, in 
     
          modelfit(xgb1, train, predictors)  File "
      
       ", line 22, in modelfit    feat_imp = pd.Series(alg.booster().get_fscore()).sort_values(ascending=False)TypeError: 'str' object is not callable解决方案：    使用 “alg.get_booster().get_fscore()”  代替 “alg.booster().get_fscore()”.