Matplotlib Pandas:在堆叠的条形图中显示列名称 [英] Matplotlib Pandas : display columns name inside a stacked barchart

问题描述

给出一个熊猫数据框:

                    response_min                                     \
device_mac                   0dd8d       1f3cc       61ff6      623ce   
datetime                                                                
2016-05-25 08:00:00       0.000000    4.250000    0.250000   0.000000   
2016-05-25 12:00:00       0.000000    0.000000    0.000000   0.000000   
2016-05-25 16:00:00       0.000000    0.000000    0.000000   0.000000   
2016-05-26 08:00:00      12.133333  119.666667    0.250000   0.000000 

这样做:

ax = df.plot(kind='bar',stacked=True,logy=True)

for p in ax.patches:
    ax.annotate(str(p.get_height()), (p.get_x() * 1.005, p.get_height() * 1.005))

我得到了:

如何继续在栏中显示列名而不是值?

非常感谢.

(剧透警报)在斯坦利的建议和调试器的帮助下，我设法做到了这一点(注意:对数刻度已关闭):

(Spoiler alert) Following Stanley's advice and the help of the debugger i managed to do it like this (NB: log scale is off):

ax = df.plot(kind='bar',stacked=True,legend=True,width=0.2)


# define a letter for each column
list_column =[ (chr(idx + ord('a')),column) for idx,column in enumerate(list(df.columns.values))]


# set the position of the annotation for each container
def set_text_position(container):
    xy_pos  = container._rect_transform._a._boxout._points
    y_start = xy_pos[0][1]
    y_end = xy_pos[1][1]
    if y_end-y_start>0:
        y_text = (y_start + ((y_end - y_start) / 2))
        return y_text
    else:
        return 0

def set_legend(ax):
    for j, column in enumerate(list_column):
        ax.legend_.texts[j]._text = list_column[j][0] + " :" + list_column[j][1]


def annotate_bar(ax, i):
    for j, column in enumerate(list_column):
        height = set_text_position(ax.containers[j][i])
        if height > 0:
            ax.annotate(column[0], xy=(i, height), xycoords="data",
                  xytext=(i+0.5, height), va="center", ha="center",
                        bbox=dict(boxstyle="square", fc="w"),arrowprops=dict(arrowstyle="-"))


for i, label in enumerate(list(df.index)):
    annotate_bar(ax, i)

ax.legend(loc='upper center', bbox_to_anchor=(0.5, 1.3), ncol=3, fancybox=True, shadow=True)
set_legend(ax)
ax.margins( 1, None )
ax.set_ylabel('dwell time (minutes)')
ax.set_xlabel("")

推荐答案

类似此处的内容:

import pandas as pd
import matplotlib.pyplot as plt

df=pd.DataFrame({'Devices': [ 'dev1', 'dev2', 'dev3', 'dev4'],
                 'Response': [4.25,0.,119.,12.1],})

df = df.set_index('Devices')
df.plot(kind='bar',  title='Response')
ax = df.plot(kind='bar',  title='Response')
ax.set_ylim(0, 130)
for i, label in enumerate(list(df.index)):
    height = df.ix[label]['Response']
    ax.annotate(str(label), (i, height + 0.3))

plt.show()

这篇关于Matplotlib Pandas:在堆叠的条形图中显示列名称的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！