如何在python中转换月份列制作条形图？

Question

我有这样的数据框。月份列是字符串类型。 我想制作201501到201505的条形图，x轴是月，而y轴是total_gmv。 x格式就像2015年2月2015年1月。那么如何使用python实现它呢？感谢。

month   total_gmv
201501  NaN
201502  2.824294e+09
201503  7.742665e+09
201504  2.024132e+10
201505  6.705012e+10

Answer 1

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

df = pd.DataFrame(
    {'month': ['201501', '201502', '201503', '201504', '201505'], 
     'total_gmv': [np.nan, 2.824294e+09, 7.742665e+09, 2.024132e+10, 6.705012e+10]})

df['month'] = pd.to_datetime(df['month'], format='%Y%m').dt.month
df = df.set_index('month')

print df
df.plot(kind='bar')
plt.show()

结果：

          total_gmv
month              
1               NaN
2      2.824294e+09
3      7.742665e+09
4      2.024132e+10
5      6.705012e+10

Answer 2

你应该能够强制月份为时间戳，然后将其设置为索引并绘制它。

df['month'] = pd.to_datetime(df.month)
ax = df.set_index('month').plot(kind='bar')

您可能需要更改日期格式。

import matplotlib.dates as mdates
ax.xaxis.set_major_formatter= mdates.DateFormatter('%b, %Y')

检查here for more

Answer 3

您应该使用matplotlib.pyplot和calendar模块。

import matplotlib.pyplot as plt
import calendar

#change the numeric representation to texts (201501 -> Jan,2015)
df['month_name'] = [','.join([calendar.month_name[int(date[-1:-3]),date[-3:]] for date in df['month']

#change the type of df['month'] to int so plt can read it
df['month'].apply(int)

x = df['month']
y = df['total_gmv']
plt.bar(x, y, align = 'center')

#i'm not sure if you have to change the Series to a list; do whatever works
plt.xticks =(x, df['month_name']) 
plt.show()

Answer 4

之前的回复有一些线索，但它没有显示详尽的答案。 您必须设置自定义xtick标签并像这样旋转它：

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

df = pd.DataFrame(
    {'month': ['201501', '201502', '201503', '201504', '201505'], 
     'total_gmv': [np.nan, 2.824294e+09, 7.742665e+09, 2.024132e+10, 6.705012e+10]})
df['month'] = pd.to_datetime(df['month'], format='%Y%m', errors='ignore')

ax = df.plot(kind='bar')
ax.set_xticklabels(df['month'].dt.strftime('%b, %Y'))
plt.xticks(rotation=0)
plt.show()