sample.py

        try:
            df_a = pd.merge(x, y)
        except pd.errors.MergeError: # occurs if only interventionids are used
            df_a = None
        try:
            df_b = pd.merge(x, y)
        except pd.errors.MergeError: # occurs if only attributeids are used
            df_b = None
            
        if df_a is None and df_b is None:
            logging.error("Problems with merging!")
        elif df_a is None:
            df = df_b
        elif df_b is None:
            df = df_a
        else:
            df = pd.concat([df_a, df_b])