Mapping values from series over a column to replace nan values pandas

Question

I have a DataFrame which has job numbers and the customer names associated with that job. There are instances where the job numbers have no customer name and therefore is null. I have a separate series which has these job numbers as index and the missing customer names to replace the null values, based on the job numbers. I am not entirely sure how I would go about mapping this over the original DataFrame column.

This is the original DataFrame (df):

   Job Number  Customer 
0    02123      Paul F
1    46456      nan
2    56823      Kevin T
3    62948      nan

The series to replace the nan values:

Job Number
  46456     Kara L
  62948     Sabrina M
  Name: Customers, dtype: object

The final output I need is:

   Job Number  Customer 
0    02123      Paul F
1    46456      Kara L
2    56823      Kevin T
3    62948      Sabrina M

I hope this makes sense. I have had a look at other answers such as using: df['Customer'] = df['Job Number'].map(customers) but that didn't work or test['Customer'] = df['Customer'].combine_first(df['Customer'].map(customers)) .

I wasn't sure how to paste code into here so I have written out the df and series by hand.

Any help would be greatly appreciated.

Answer 1

You could use reset_index with combine_first :

(df.set_index('JobNumber').squeeze()
   .combine_first(customers.set_index('Job').squeeze())
   .reset_index())

       index  Customer
    0   2123     Paul F
    1  46456     Kara L
    2  56823    Kevin T
    3  62948  Sabrina M

Answer 2

Here is necessary use map by column Job Number instead Customer :

df['Customer'] = df['Customer'].combine_first(df['Job Number'].map(customers))
print (df)
   Job Number   Customer
0        2123     Paul F
1       46456     Kara L
2       56823    Kevin T
3       62948  Sabrina M

Detail :

print (df['Job Number'].map(customers))
0          NaN
1       Kara L
2          NaN
3    Sabrina M
Name: Job Number, dtype: object