How to calculate percentage for row counts in groupby in Python and generate bar plot-CodePudding

I want to calculate the percentage, for each id and generate a bar plot.

Here an example of my data:

id  LGA        Status
1   Banyule     Referred
2   Hepburn     Referred
3   Kingston    Not Referred
4   Darebin     Not Referred
5   Darebin     Managed Externally
6   Darebin     Managed Externally
7   Mansfield   Managed Externally
8   Casey       Referred
9   Mitchell    Referred
10  Mitchell    Not Referred
11  Moreland    Referred
12  Whittlesea  Not Referred
13  Glen Eira   Not Referred
14   Dandenong  Referred
15  Hume        Not Referred
16  Hume        Managed Externally
17  Campaspe    Not Referred
18  Melbourne   Not Referred
19  Melbourne   Not Referred

I've used "groupby" function to calculate counts for column "LGA" and "Status" and generate a bar plot.

Sample code;

 df['Status'].value_counts().plot(kind='bar')
 df['LGA'].value_counts().plot(kind='bar')

I am not sure how to do it elegantly if I am interested in plotting percentages for same columns and generate bar plot separately.

Expected output: I have derived below output using excel

Status             % of Grand Total
Not Referred         58.42%
Referred             23.68%
Managed Externally   17.89%
Grand Total          100.00%

Expected Bar Chart:

Any help would be appreciated.

CodePudding user response：

I believe this is what you are looking for:

temp_df = (df.groupby('Status').size().sort_values(ascending=False) / df.groupby('Status').size().sort_values(ascending=False).sum())*100
    
ax = temp_df.plot(kind='bar')
    
ax.bar_label(ax.containers[0])
        
plt.show()