meadover
commented
4 years ago Sergio's -estfe- command syntax is based on the syntax for Stata's -estimates table- and Ben Jann's -esttab- command. In the syntax for these two classic commands, a period appearing in the list of estimation results is interpreted as a reference to the most recent set of estimation results in memory.
You write:
estfe . est1 est2, labels(year "Year FE" industry "Industry FE" municipio "Municipality FE")
Note that you have inserted a period between the command name and the name of the first set of estimation results, -est1-. So my guess is that, by including that period, you are indeed asking for -est3-. Since-reghdfe- did not complete its work, I suppose it erased your -est2- results and never stored any -est3- results. Hence -estfe- correctly reports:
last estimation results not found, nothing to store
r(301);Hope this helps.
Mead
giofsantos11
Behavior
I am writing code to be run by a data center with confidential data. I wan to run several regressions and produce the output. I am not sure if all specifications will run without problem but would like to have the output for the ones that do. Thus I am capturing the errors so that the code does not stop. The error is occurring when I estfe after the estimation.
For example, I run three regressions, store the estimation results and the third cannot be run due to insufficient observations. I capture the error so the code continues to run. eststo: quietly reg output win, absorb(year) eststo: quietly reghdfe output win, absorb(industry year) capture eststo: quietly reghdfe output win, absorb(municipio year) And then write and get error: estfe . est1 est2, labels(year "Year FE" industry "Industry FE" municipio "Municipality FE") last estimation results not found, nothing to store r(301);
Why is this happening even though I am not asking for est3 to be used?