Python pandas - 将 groupby output 写入文件

Question

I used the following to get proportion information on my data:

>>>testfile = pd.read_csv('CCCC_output_all_FINAL.txt', delimiter="\t", header=0)
>>> testdf = pd.DataFrame({'Proportion': testfile.groupby(('Name','Chr','Position','State')).size() / 39})
>>> testdf.head(5)
                                        Proportion
Name    Chr Position  State           
S-3AAAA 16  27557749  4        0.025641
                                    5        0.076923
                                    6        0.025641
S-3AAAC 15  35061490  2        0.076923
                                    4        0.025641

>>> testdf.to_csv('CCCC_output_summary.txt', sep='\t', header=True, index=False)

The output file only has the column Proportion . I'd like the following table output:

Name    Chr    Position     State     Proportion
S-3AAAA  16     27557749     4         0.025641
S-3AAAA  16     27557749     5         0.076923
S-3AAAA  16     27557749     6         0.025641
S-3AAAC  15     35061490     2         0.076923
S-3AAAC  15     35061490     4         0.025641

Is it possible/easy to write the pandas output to a file like this?

Answer 1

使用reset_index() ：

testdf.reset_index().to_csv('CCCC_output_summary.txt', sep='\t', header=True, index=False)

Answer 2

Recently, I had to work with an Excel file that has 2 columns, with headers 'Dog Breed' and 'Dog Name'. I came up with the following code (tested with Python 3.11.0 ) that uses groupby() and prints the grouped data into a .csv file.

from pathlib import Path
import pandas as pd

p = Path(__file__).with_name('data.xlsx')
q = Path(__file__).with_name('data-grouped.csv')

df = pd.read_excel(p)
groups = df.groupby('Dog Breed', sort=False)

with q.open('w') as foutput:
for g in groups: # For each group
    foutput.write(f"{g[0]}, {len(g[1])}") # Record the number of dogs in each group
    for e, (index, row) in enumerate(g[1].iterrows()): # Iterating over the group's dataframe
        name = str(row['Dog Name'])
        if(e == 0):
            mystr = f",{name}\n"
        else:
            mystr = f",,{name}\n"
        foutput.write(mystr)

data.xlsx:

data-grouped.csv:

Answer 3

I had the same problem. reset_index() as explained above did not work for me. I used an answer from another Stackoverflow and it worked wonderfully. Details are below.

Input csv has data under following two columns: Item Code, Quantity

Output needed: Average quantity grouped by item and both columns to be part of csv.

Initial code:

import pandas as pd

data_directory = os.path.join("D:\\data")
df = pd.read_csv(os.path.join(data_directory, "input_file.csv"))

df_avg = df.groupby("Item Code")["Quantity"].mean()
df_avg.reset_index().to_csv(os.path.join(data_directory,'output_file.csv'), sep='\t', header=True, index=False )

Output received: Only the average quantity was written to output file

Following code solved the problem:

import pandas as pd

data_directory = os.path.join("D:\\data")
df = pd.read_csv(os.path.join(data_directory, "input_file.csv"))

df.groupby("Item Code")["Quantity"].mean().reset_index()[["Item Code", "Quantity"]].to_csv(os.path.join(data_directory,'output_file.csv'))

By the above code, I got the output file which has two columns: Item Code and Quantity and the second column contains average of quantity for each Item code.

Other stack overflow reference: Pandas groupby to to_csv