#clf.fit(X_total_train,y_total_train, sample_weight=w_total_train)
mse = mean_squared_error(y_total_test, clf.predict(X_total_test))
print("MSE: %.4f" % mse)
#clf.evals_result()
print clf.score(X_total_train,y_total_train)
#print clf.best_params_
joblib.dump(clf, os.path.expanduser('/eos/user/i/ivovtin/HHggbb/HHbbgg_ETH_devel/Training/plots/optimization/simlple_Test_binary_st.pkl'), compress=9)
import matplotlib.pyplot as plt
from xgboost import plot_tree
from sklearn.metrics import accuracy_score
#plotting.plot_input_variables(X_sig,X_bkg,branch_names)
#plt.show()
plotting.plot_classifier_output(clf,X_total_train,X_total_test,y_total_train,y_total_test,outString="xbrg_test_st_values_notcut")
#plt.show()
#fpr,tpr = plotting.plot_roc_curve(X_total_train,y_total_train,clf)
#plotting.print_roc_report(fpr,tpr)
#plt.savefig(utils.IO.plotFolder+"ROC_train.eps")
#plt.show()
#fpr,tpr = plotting.plot_roc_curve(X_total_test,y_total_test,clf)
#plotting.print_roc_report(fpr,tpr)
#plt.show()
plt.bar(range(len(clf.feature_importances_)), clf.feature_importances_)
plt.savefig(utils.IO.plotFolder+"importance1.eps")
#plt.show()
#
xgb.plot_importance(clf)
import matplotlib.pyplot as plt
reload(plt)
#outTag = '2018/dev_legecy_runII_Mjj_woMjjcut_v2/'
outTag = '2018/dev_legecy_runII_ext_rho_rew_v3/'
joblib.dump(clf,
os.path.expanduser(utils.IO.plotFolder + outTag +
'simlple_Test_binary_st.pkl'),
compress=9)
#plotting.plot_input_variables(X_sig,X_bkg,branch_names)
#plt.show()
plotting.plot_classifier_output(clf,
X_total_train,
X_total_test,
y_total_train,
y_total_test,
outString=utils.IO.plotFolder + outTag +
"classifierOutputPlot_xbrg_test_st_values")
#plt.show()
#fpr,tpr = plotting.plot_roc_curve(X_total_train,y_total_train,clf)
#plotting.print_roc_report(fpr,tpr)
#plt.savefig(utils.IO.plotFolder+"ROC_train.eps")
#plt.show()
#fpr,tpr = plotting.plot_roc_curve(X_total_test,y_total_test,clf)
#plotting.print_roc_report(fpr,tpr)
#plt.show()
#plt.bar(range(len(clf.feature_importances_)), clf.feature_importances_)
#plt.savefig(utils.IO.plotFolder+outTag+"importance1.eps")
outTag = mass_range + 'mass'
folder = str(pklfolder) + '_' + str(sig) + '_' + outTag + '_' + str(year)
if not os.path.exists(folder):
os.mkdir(folder)
joblib.dump(
clf,
os.path.expanduser(str(folder) + '/' + outTag + '_XGB_training_file.pkl'),
compress=9)
#plotting.plot_input_variables(X_sig,X_bkg,branch_names)
#plt.show()
plotting.plot_classifier_output(clf,
X_total_train,
X_total_test,
y_total_train,
y_total_test,
outString=str(folder) + '/' + outTag +
"_classifierOutputPlot_xbrg_test_st_values")
plt.clf()
#plt.show()
#fpr,tpr = plotting.plot_roc_curve(X_total_train,y_total_train,clf)
#plotting.print_roc_report(fpr,tpr)
#plt.savefig(utils.IO.plotFolder+"ROC_train.eps")
#plt.show()
#fpr,tpr = plotting.plot_roc_curve(X_total_test,y_total_test,clf)
#plotting.print_roc_report(fpr,tpr)
#plt.show()
#plt.bar(range(len(clf.feature_importances_)), clf.feature_importances_)