How to concatenate values of all rows in a dataframe into a single row without altering the columns?

Question

I have a data frame input that looks like:

  col1 col2 col3
0    3    1  NaN
1  NaN    7    8

How to collapse all rows while concatenating the data in the rows with ', '?

The desired data frame output:

  col1  col2 col3
0    3  1, 7    8

---

Sample input code:

import pandas as pd
import numpy as np


d = {'col1': ["3", np.nan], 'col2': ["1", "7"], 'col3': [np.nan, "8"]}
df = pd.DataFrame(data=d)

Answer 1

agg + dropna + str.join comes to mind.

df.agg(lambda x: ', '.join(x.dropna())).to_frame().T

  col1  col2 col3
0    3  1, 7    8

There are other solutions, my peers will find them for you :)

Answer 2

pd.DataFrame(
    [[
        ', '.join(map(str, map(int, filter(pd.notna, c))))
        for c in zip(*df.values)
    ]], columns=df.columns
)

  col1  col2 col3
0    3  1, 7    8

Answer 3

One way to get what you want would be to create a new dataframe with the same columns as your old dataframe, and populate the first index with your desired data. In your case, your desired data would be a list of each column, joined by ', ', and with your NaN values removed:

new_df = pd.DataFrame(columns=df.columns)

for col in df.columns:
    new_df.loc[0, col] = ', '.join(df[col].dropna().tolist())

>>> new_df
  col1  col2 col3
0    3  1, 7    8