plot_importance多分类、排序不匹配、图片数值不显示

多分类

params = {'booster': 'gbtree',
		  'objective': 'reg:squarederror',
		  "learning_rate": 0.01,
		  'n_estimators': 500,
		  "missing": -1}

multioutputregressor_xgb = MultiOutputRegressor(xgb.XGBRegressor(**params)).fit(train_x, train_y)
predict_values = multioutputregressor_xgb.predict(test_x)

for items in multioutputregressor_xgb.estimators_:
	items.get_booster().feature_names = x_columns_list
	empty_dict = {k:round(v,4) for k,v in items.get_booster().get_score(importance_type='gain').items()}
	if bool(empty_dict):
		plot_importance(empty_dict, max_num_features=10, importance_type='gain',show_values=True, title="Feature importance")

	imp["name"] = items.get_booster().feature_names
	imp[y_columns_list[i]] = items.feature_importances_
	imp.round(6)
	i+=1
plt.show()
imp.to_csv(importance_path, sep="\t", float_format="%.6f")

排序不匹配

model.feature_importances_的重要性排名默认使用gain，而xgb.plot_importance默认使用weight，所以:

xgb.plot_importance(model,max_num_features=10,importance_type='gain')

图片数值不显示

打开plotting.py 修改：

def plot_importance(booster, ax=None, height=0.2,
                    xlim=None, ylim=None, title='Feature importance',
                    xlabel='F score', ylabel='Features', fmap='',
                    importance_type='weight', max_num_features=None,
                    grid=True, show_values=True, **kwargs):

增加max_digits=3,,修改show_values，原来是这样：

if show_values is True:
    for x, y in zip(values, ylocs):
        ax.text(x + 1, y, x, va='center')

更改成：

if max_digits is not None:
	t = values
	lst = list(t)
	if len(str(lst[0]).split('.')[-1]) > max_digits:
		values_displayed = tuple([('{:.' + str(max_digits) + 'f}').format(x) for x in lst])
	else:
		values_displayed = values

if show_values is True:
	for x, x2, y in zip(values, values_displayed, ylocs):
		dx = np.max(values) / 100
		ax.text(x + dx, y, x2, va='center')