Pandas万花筒：让绘图变得更美观

output_notebook()              p1 = data.plot_bokeh.scatter(x='Hue',                                                y='Proline',                                                category='class',                                                title='Proline and Hue by wine class',                                                show_figure=False)                                        p2 = data[['Hue','class']].groupby(['class']).mean().plot.bar(title='Mean Hue per Class')              df_hue = pd.DataFrame({                      'class_1': data[data['class'] =='1']['Hue'],                      'class_2': data[data['class'] =='2']['Hue'],                      'class_3': data[data['class'] =='3']['Hue']},                      columns=['class_1', 'class_2', 'class_3'])              p3 = df_hue.plot_bokeh.hist(title='Distribution perClass: Hue')              df_proline = pd.DataFrame({                      'class_1': data[data['class'] =='1']['Proline'],                      'class_2': data[data['class'] =='2']['Proline'],                      'class_3': data[data['class'] =='3']['Proline']},                      columns=['class_1', 'class_2', 'class_3'])              p4 =df_proline.plot_bokeh.hist(title='Distribution per Class: Proline')              pandas_bokeh.plot_grid([[p1, p2],                                          [p3, p4]], plot_width=450) 

为内置的Pandas绘图功能添加多个第三方后端，这大大增强了该库用于数据可视化的能力。从此之后，pandas就可以集美貌与实用于一身啦。

（编辑：晋中站长网）

【声明】本站内容均来自网络，其相关言论仅代表作者个人观点，不代表本站立场。若无意侵犯到您的权利，请及时与联系站长删除相关内容!