Home » excel » python – Working with Pandas DataFrame / Sorting

python – Working with Pandas DataFrame / Sorting

Posted by: admin April 23, 2020 Leave a comment


I’m working with a big dataset within Excel in which I’m trying to sort a number by top 25 per index value.

The datasite looks like this:


The Final PAC ID is the company number and changes (this does not show in the given data). The PAC contribution is the number I want to sort by.

So for example, there will be 50 contributions done by company C00003590, to different candidates with amount ‘PAC contribution’, I would like to sort the top 25 contributions done per company.

I’ve tried working with dictionaries, creating a dictionary for each company and adding in the candidate numbers as a string key, and the contribution as a value.

The code I have so far is the following (this might be the completely wrong way to go about it though):

import pandas as pd

df1 = pd.read_excel('Test2.xlsx')

dict_company = {}
k1 = str(df1['Final PAC ID'])
k2 = str(df1['Candidate ID'])

for each in range(0,100):
    dict_company[k1)[each]] = {}
    dict_company[k1)[each]] = k2[each]
    if each % 50 == 0:


for each in range(0,100):
    dict_company[k1][k2][each] = round(float(k1[each]))
    if each % 50:

How to&Answers:

I think you need nlargest:

df1 = df.groupby('Final PAC ID')['PAC contribution'].nlargest(50)

If need all columns:

cols = df.columns[~df.columns.isin(['PAC contribution','Final PAC ID'])].tolist()
df1 = df.set_index(cols)
         .groupby('Final PAC ID')['PAC contribution']

Another solution (can be slowier):

df1 = df.sort_values('PAC contribution', ascending=False).groupby('Final PAC ID').head(50)

Last save to excel by to_excel:



df.groupby('Final PAC ID').head(50).reset_index(drop=True)


You can use groupby in conjunction with a dictionary comprehension here. The result is a dictionary containing your company names as keys and the sub dataframes with top 25 payments as values:

def aggregate(sub_df):
    return sub_df.sort_values('PAC contribution', ascending=False).head(25)

grouped = df.groupby('Final PAC ID')
results = {company: aggregate(sub_df)
           for company, sub_df in grouped}