Problems with regression modus

[ylnguyen]

The network is failing when the modus is set to regression and coarse is set to False. I get the following error:
ValueError: Input contains NaN, infinity or a value too large for dtype('float32').
When I trace this error, it seems to be related to the XGBRegressor from the XGBoost library. For now, I just manually add the estimators and leave the XGBRegressor out of the ensemble:
estimators = ['SVR', 'RFR', 'ElasticNet', 'Lasso', 'AdaBoostRegressor', 'LinR', 'Ridge']
if modus == 'binary_classification': experiment.binary_classification(coarse=coarse)
elif modus == 'regression': experiment.regression(estimators=estimators, coarse=None)
elif modus == 'multiclass_classification': experiment.multiclass_classification(coarse=coarse)