Take df_1:
df_1 = pd.DataFrame({'date':['2021-01-25','2021-01-26','2021-01-27'],
'a':[np.nan,np.nan,np.nan]})
date a
0 2021-01-25 NaN
1 2021-01-26 NaN
2 2021-01-27 NaN
And take df_2:
df_2 = pd.DataFrame({'date':['2021-01-27','2021-01-28','2021-01-29'],
'b':[1,2,3]})
date b
0 2021-01-27 1
1 2021-01-28 2
2 2021-01-29 3
I need to fill column a value getting column b value, if they have the same dates.
You may merge the two dataframes based on date and extract the desired column from second dataframe. Like this:
df_1.merge(df_2,how='left', on='date')[['date', 'b']].rename(columns={'b': 'a'})
Try this:
d = df_2.set_index('date').squeeze().to_dict()
df_1['a'] = df_1['a'].fillna(df_1['date'].map(d))
You can set 'date' as the index in both data frames and fill missing values using the method fillna:
df_1.set_index('date')['a'].fillna(df_2.set_index('date')['b']).reset_index()
Output:
date a
0 2021-01-25 NaN
1 2021-01-26 NaN
2 2021-01-27 1.0
I think you should use pd.merge:
pd.merge(df_1['date'],df_2, on='date', how='outer').dropna(axis = 0)