implemented hard and soft voting option for prediction of ensemble classifiers in case of predict_prob=False
added test for multi target classification ensemble, adapted test of single target classification ensemble
@ppdebreuck let me know if you need anything else(:
Also I want to note that with the current changes, the shapes of df_mean and df_std differ for soft voting (df_std contains stds on class probability predictions, df_mean integer class labels), but I don't know if anything else would make more sense.
Resolves #225 .
predict_prob=False@ppdebreuck let me know if you need anything else(:
Also I want to note that with the current changes, the shapes of
df_meananddf_stddiffer for soft voting (df_stdcontains stds on class probability predictions,df_meaninteger class labels), but I don't know if anything else would make more sense.