Merging multiple rows into one in dataframe

Question

I have a dataframe which looks like:

df = 
 |Name Nationality Family etc.....
0|John Born in Spain. Wife
1|nan But live in England son
2|nan nan daughter

Some columns only have one row but others have multiple answers over a few rows, how could i merge the rows onto each other so it would look something like the below:

df = 
 |Name Nationality Family etc....
0|John Born in Spain. But live in England Wife Son Daughter

• Umar.H: whats the logic of join? pretty simple to do. Assuming any nan values under Name mean it belongs to most recent value? if so df['Name'].ffill().groupby('Name').agg(' '.join) should do the trick - you may need to specify your columns and join types if you have numbers,dates etc.

Answer 1

Perhaps this will do it for you:

import pandas as pd
# your dataframe
df = pd.DataFrame(
 {'Name': ['John', np.nan, np.nan],
 'Nationality': ['Born in Spain.', 'But live in England', np.nan],
 'Family': ['Wife', 'son', 'daughter']})
def squeeze_df(df):
 new_df = {}
 for col in df.columns:
 new_df[col] = [df[col].str.cat(sep=' ')]
 return pd.DataFrame(new_df)
squeeze_df(df)
# >> out: 
# Name Nationality Family
# 0 John Born in Spain. But live in England Wife son daughter

I made the assumption that you only need to do this for one single person (i.e. squeezing/joining the rows of the dataframe into a single row). Also, what does "etc...." mean? For example, will you have integer or floating point values in the dataframe?