calculate growth rate in panel data with time series

[Alan Walker]

Hi,

i have a panel data with country names, country ID, years etc
i need to compute growth rate of a variable let say X. in order to do so i first computed this variable X. then i changed it to log form by

gen l_X= ln(X)

then when i tried to get growth rate i used

gen grw_X= l_X - l_X[_n-1]

it works but when i analysed outcome i noticed that time series are not accounted and when i have panel data with years lets say 2000-2010 and multiple different countries with different years, those years are inconsistent for ex. and when growth rate is calculated there is a significant change from end of last year in USA and first year in China as the formula computes previous year of different country and not start from a country year one, it just taking into consideration previous without accounting that it is different country already. I hope that makes sense.

year country

2000 usa
2001 usa
2002 usa
2003 usa
2004 usa
2005 usa
2000 china
2001 china
2002 china
2003 china
2000 india
.
.
.

i would be most grateful if someone could help with this problem, thanks in advance

[Nick Cox]

https://www.statalist.org/forums/for...out-if-command covers the question you have. (May not seem so from the thread title, but I think it does.)

[Alan Walker]

thank you so much Nick Cox