Fill NA values in dataframe by mapping a double indexed groupby object

Question

I have a data frame of bank customer details and I have created a double indexed group by object grouped by income type and education returning the means of the actual income of those customers, I created an example below.

I'm looking to remap the mean income of this group by object back into the data frame to fill in NA values in the income column.

I can map with a single index but I'm not sure how to apply a doubled indexed map which provides a better assessment of averages. Any help is appreciated, thanks.

income_means = df.groupby(["income_type", "education"])['total_income'].agg('mean')



                                  total_income
                                      mean
income_type    education                  
               bachelors              50456
professional   high-school            40657
               elementary             30576

               bachelors              55069
management     high-school            50687
               elementary             40687


df['total_income'] = df['total_income'].fillna(map(income_means))