Replacing the variable's missing observations with the existing mean of that variable

[Nick Baradar]

Hello Everyone,
I have data on income and I would like to replace the variable's missing observations with the existing mean of that variable itself. The command that I am using now comprises three lines. I would highly appreciate if you could advise me on whether I could write in that command in a single line, more efficiently. Thank you very much in advance!

The command that I have now is the following:

     egen income_mean=mean(income_recoded)
     replace income_recoded = income_mean if income_recoded==.
     drop income_mean

And below please find that variable:

* Example generated by -dataex-. To install: ssc install dataex
clear
input float income_recoded
 .75
2.75
2.25
2.25
2.75
2.75
2.25
 .25
4.75
1.75
4.25
3.25
1.75
6.25
6.25
 .75
3.25
2.25
2.25
2.25
3.75
4.75
3.75
3.75
3.25
1.75
4.25
3.25
1.25
1.75
3.25
3.75
3.25
3.25
1.25
1.75
2.25
4.75
2.75
2.75
4.25
3.75
3.75
3.75
2.25
3.25
3.75
4.75
1.25
6.25
6.25
3.25
6.25
8.75
 .75
2.25
4.25
3.75
4.75
4.25
6.25
3.75
1.75
1.75
2.75
4.25
4.75
3.75
1.75
2.25
3.75
2.25
3.75
4.25
1.25
3.25
2.75
2.75
3.75
2.25
6.25
   .
   .
3.75
1.75
1.25
1.75
 .75
3.25
2.25
2.75
6.25
   .
2.75
2.75
2.75
6.25
   .
1.75
2.25
end

[Nick Cox]

The same effect is achieved with

su income_recoded, meanonly 
replace income_recoded = r(mean) if income_recoded == .

without bothering with a new variable. The bigger deal is that this would be widely reckoned to be the simplest and also the most problematic imputation method imaginable, but I guess you know that -- not to mention that mean income is an unstable summary.