将交叉表熊猫绘制成饼图

Question

I have a dataframe that i run a crosstab function and i want to plot the crosstab table into a pie chart .

The problem is in the value and label parameters in the pie chart.

dataframe :

    event_type    date      event_mohafaza  number_person   groups
0   watch movie  2020-08-14     loc1             25         group1
1   stay at home 2020-08-14     loc3             32         group1
2   watch movie  2020-08-14     loc2             45         group2
3   swimming     2020-08-14     loc4             12         group1
4   stay at home 2020-08-14     loc2             45         group3
5   watch movie  2019-06-14     loc5             22         group1
6   watch movie  2020-08-14     loc2             10         group4
7   watch movie  2020-08-14     loc1             44         group1
8   stay at home 2020-08-14     loc3             22         group3
9   swimming     2020-08-14     loc2             32         group2
10  watch movie  2019-09-14     loc1             17         group1
11  camping      2020-08-14     loc4             27         group1
12  watch movie  2020-08-14     loc5             43         group3
13  meeting      2019-06-14     loc2             33         group2
14  camping      2020-08-14     loc1             21         group4

crosstab tabel:

event_type  camping        camping         meeting         stay at home    \
year_month                                                                  
2019/06                 0               0               1               0   
2019/09                 0               0               0               0   
2020/08                 1               1               0               3   

event_type  swimming       swimming        watch movie     
year_month                                                 
2019/06                 0               0               1  
2019/09                 0               0               1  
2020/08                 1               1               5   

pie chart plotting:

ddf['year_month'] = ddf['date'].map(lambda x: x.strftime('%Y/%m'))
ct = pd.crosstab(ddf['year_month'], ddf['event_type'])
print(ct)


ct = ct.reset_index()
ct['labels'] = ct['year_month'] + ' ' + ct['event_type']

trace = go.Pie(labels=ct['labels'], 
hoverinfo='label+percent', 
values=ct['event_type'],  
textposition='outside',                
rotation=90)
layout = go.Layout(
        title="Percentage of events",
        font=dict(family='Arial', size=12, color='#909090'),
        legend=dict(x=0.9, y=0.5)
        )
data = [trace]
print(data)
fig = go.Figure(data=data, layout=layout)
fig.show()

Answer 1

You can draw a graph in a vertical format without changing the data frame to a horizontal format. The code format has been rewritten according to the official reference. Is this answer what you intend to do?

ddf['date'] = pd.to_datetime(ddf['date'])
ddf['year_month'] = ddf['date'].map(lambda x: x.strftime('%Y/%m'))

ddf['labels'] = ddf['year_month'] + ' ' + ddf['event_type']

import plotly.graph_objects as go

fig = go.Figure(data=[go.Pie(labels=ddf['labels'],
                             hoverinfo='label+percent',
                             values=ddf['number_person'], 
                             textposition='outside', 
                             rotation=90)])

fig.update_layout(title="Percentage of events",
                  font=dict(family='Arial', size=12, color='#909090'),
                  legend=dict(x=0.9, y=0.5)
        )

fig.show()