Python熊猫-分组，计数，绘制多系列

Question

我从一项调查中得到以下数据集，该数据集为每个参与者提供了一份食物清单，并请他们对他们本周食用食物的可能性进行排名。我想在图表上画出每种食物的每种可能性的计数。

Person    Food     Label
John      Pizza    Likely
John      Chinese  Unlikely
John      French   Very Unlikely
Debbie    Pizza    Unlikely
Debbie    Chinese  Very Likely
Debbie    French   Very Unlikely

例如：

Pizza     Likely         1
Pizza     Unlikely       1
Chinese   Unlikely       1
Chinese   Very Unlikely  1
French    Very Unlikely  2

到目前为止，我已经将文件读入数据帧并进行了一些基本清理。

import pandas as pd

raw_data = pd.read_excel('my_file_path')

#cleaning code
clean_data = raw_data(clean)

results = clean_data.groupby(['Food', 'Label']).count()

Answer 1

我相信您需要在Person之后添加列groupby，用unstack重塑形状，并用DataFrame.plot.bar绘制：

results = clean_data.groupby(['Food', 'Label'])['Person'].count().unstack(fill_value=0)

使用crosstab的另一种解决方案：

results = pd.crosstab(clean_data['Food'], clean_data['Label'])

print (results)
Label    Likely  Unlikely  Very Likely  Very Unlikely
Food                                                 
Chinese       0         1            1              0
French        0         0            0              2
Pizza         1         1            0              0

results.plot.bar()