ef29cd9bde8695645d63ae7f568fb298809e7b19,tpot/builtins/dataset_selector.py,DatasetSelector,fit,#DatasetSelector#,43

Before Change

        self: object
            Returns a copy of the estimator
        
        subset_files = os.listdir(self.subset_dir)
        self.subset_i = self.subset_dir + "/" + self.sel_subset_fname
        features_i_df = pd.read_csv(self.subset_i, sep="\t", header=0)
        if isinstance(X, pd.DataFrame): // use columns" names
            self.feature_names = list(X.columns.values)

After Change

        if isinstance(self.sel_subset, int):
            self.sel_subset_name = subset_df.index[self.sel_subset]
        elif isinstance(self.sel_subset, list):
            self.sel_subset_name = []
            for s in self.sel_subset:
                if isinstance(s, int):
                    self.sel_subset_name.append(subset_df.index[s])
                else:
                    self.sel_subset_name.append(s)
        else: // self.sel_subset is a string
            self.sel_subset_name = self.sel_subset

        sel_features = subset_df.loc[self.sel_subset_name, "Features"]
        if not isinstance(sel_features, str):
            sel_features = ";".join(sel_features.tolist())

        sel_uniq_features = set(sel_features.split(""))

Instances