reghdfe how to get e(clust#)

[haiyan lin]

Hi all,

Greetings. I am writing to ask how to export the number of each cluster used in the regression after reghdfe. I have panel data with country-year pairs. In the regression, I control fixed effects at the country and year level. Since my data is unbalanced, I want to know how many countries have been used in each regression. From the help file, it seems the information is stored in e(clust#). I tried to display `e(clust1)' or `e(clustcountry)', but didn't get the information I wanted. I also checked the "ereturn list" and the e(df_a) contains the total number of clusters including both countries and years. But I want to get the number of countries separately and I didn't find any local or global vars containing such information.

Please help me if you know how to do it. Thank you very much in advance!

Regards,
Haiyan

[haiyan lin]

I solved my problem with the below code:

distinct country if e(sample)
display `r(ndistinct)'

Hope it helps.