data_handler.py 文件源码

def label_metadata(label_matrix, label_col):
    # Check whether the column value is given as index (number) or name (string) 
    try:
        label_col = int(label_col)

        # If given as number, take the name of the column out of it
        label_col = label_matrix.columns[label_col]
    except ValueError:
        pass

    # Get the unique classes in the given column, and how many of them are there
    unique_classes = pd.unique(label_matrix[label_col].ravel())

    # Map the unique n classes with a number from 0 to n
    label_map = pd.DataFrame({label_col: unique_classes, label_col+'_id':range(len(unique_classes))})

    # Replace the given column values with the mapped equivalent
    mapped_labels = label_matrix.replace(label_map[[0]].values.tolist(), label_map[[1]].values.tolist())
#     print("label_matrix", label_matrix)
#     print("mapped_labels", mapped_labels)

    # Return the mapped labels as ndarray and the label map (unique classes and number can be obtained from map)
    # np.reshape(mapped_labels[[label_col]].values, (mapped_labels.shape[0],))
    # Return the mapped labels as DataFrame and the label map (unique classes and number can be obtained from map)
    return mapped_labels[[label_col]], np.asarray(label_map) #, unique_classes, num_classes