Mean of 5 obs. in Panel Dataset

[Anass Wajib]

Hello,

I have a Panel data and I've been searching for a command allowing me to compute an average of 5 values of my "growth" variable of each country (please see blow)

The average values will be in a new column of course.


Thank you!

[Carlo Lazzaro]

Anass: please follow the FAQ when posting. Thanks.
That said, you may find what follow helpful:

. use "https://www.stata-press.com/data/r18/nlswork.dta"
(National Longitudinal Survey of Young Women, 14-24 years old in 1968)

. bysort idcode (year): egen wanted=mean(ln_wage) if year<=75


. list wanted if idcode==1

       +----------+
       |   wanted |
       |----------|
    1. | 1.525419 |
    2. | 1.525419 |
    3. | 1.525419 |
    4. | 1.525419 |
    5. | 1.525419 |
       |----------|
    6. |        . |
    7. |        . |
    8. |        . |
    9. |        . |
   10. |        . |
       |----------|
   11. |        . |
   12. |        . |
       +----------+

.

[Anass Wajib]

Carlo Lazzaro Apologies for the late response, I made sure to fully look into your link and understand it.

It worked! Thank you immensely. One last question please, since the timeline in your example starts at year 70, the command works perfect but my timeline has many missing observations (please see picture above) before the sought out years. It there a command that tells STATA to start from year X and end with year Y. Something similar as such 1986>=mean<=1990.

Thank you again.

[Nick Cox]

Stata has lots of functionality here. Check out tssmooth (official), rangestat (SSC) and commands like

egen wanted8690 = mean(cond(inrange(year, 1986, 1990), growth, .))

On the last, see Section 9 of https://journals.sagepub.com/doi/pdf...867X1101100210

The rest of the paper may be helpful.

[Anass Wajib]

Nick Cox Thank you very much for your assistance