Guangqiang Lu

I'm sorry I can't get your question idea. You want to plot the precision_recall curve, but your code is just using `plot_confusion_matrix` for confusion matrix. In fact, if you want...

@reiinakano I have added the digits parameter to control the precision for metrics function, please see the request. Hope you would check that!

#102 Here just add the digits parameters for the metrics function like plot_roc and plot_precision_recall etc. Please see the difference for comparison.

@reiinakano I have changed plot_cumulative_gain for choosing classes for ploting also with 'micro' and 'macro' as default ploting curve. You can check it out.

Because pytest of unit test for plot_cumulative_plot is not allowed for multi-classes plot originally, so this may cause pytest can't pass.

Because as we use the grid-search like param choosen method, the result is based on the train set only, not on all data, so just like the cross-validation to evaluate...

As for `plot_confusion_matrix` has one parameter that could control the size of the plot, you could change the `figsize` parameter to change the size to fit your desktop(with tuple type...

@AntonioAntovski 'plot_roc' function is based on sklearn's 'roc_curve' function, this function will check input data shape. Maybe you should not use the 'index' column, because your label is 7-classes, but...