绘制熊猫groupby对象

Question

I have a groupby object to which I applied value_counts(normalize=True) which yields the following series:

Category          Reason         
25-39 year old    Money                   0.20
                  Time                    0.10
                  Interest                0.70
40-54 year old    Money                   0.50
                  Time                    0.20
                  Interest                0.30
55+ year old      Money                   0.70
                  Time                    0.10
                  Interest                0.20

And I would like to make a bar chart, where x = category, y = values (shown above), and the "Reason" column being used as hue. See quick example in paint below:

Answer 1

Rather than using groupby() you could make a pivot table and plot it directly. Assuming your value column is called 'Value' :

from io import StringIO

import pandas as pd
import matplotlib.pyplot as plt

data = StringIO('''Category,Reason,Value
25-39 year old,Money,0.20
25-39 year old,Time,0.10
25-39 year old,Interest,0.70
40-54 year old,Money,0.50
40-54 year old,Time,0.20
40-54 year old,Interest,0.30
55+ year old,Money,0.70
55+ year old,Time,0.10
55+ year old,Interest,0.20''')

df = pd.read_csv(data)

df['Category'] = df['Category'].str.replace(' year old', '')
pvt = df.pivot_table(index='Category', columns='Reason', values='Value')
pvt.plot.bar()
plt.show()

This shows the following graph: