Difference-in-difference : did_multiplegt table - r(199)

[Benoit Decoco]

Hello,

I am doing a difference-in-difference with multiple groups and I can't seem to get a table.
My code is

did_multiplegt log_aod_mean_10km urbancode date treat, trends_nonparam(c) controls(satdum)

I get the output

When the breps option is not used, the command does not produce a table or a graph,
but the estimators requested are stored as eclass objects. Type ereturn list to see them.

but when I run

did_multiplegt log_aod_mean_10km urbancode date treat, trends_nonparam(c) controls(satdum) breps(2)

I get

command fcollapse is unrecognized
r(199);

and after this error, when I type in the console

did_multiplegt log_aod_mean_10km urbancode date treat, trends_nonparam(c) controls(satdum)

, it gives the same error.

command fcollapse is unrecognized
r(199);

I searched for fcollapse on Stata but this command does not exist, any idea how the problem could be solved?
I'd like to have a table with standard errors for the diff-and-diff estimation.

[Daniel Schaefer]

Hi Benoit, it looks like fcollapse is part of ftools. My guess is that ftools is a dependency for -did_multiplegt- and you will need to download and install ftools to get this to work.

Edit: you may also want to look through this thread: https://www.statalist.org/forums/for...large-datasets

[Benoit Decoco]

Thank you Daniel ! ftools was indeed summoned by the command. I now get the following error

variable FE1 not found

I'll look into that and if I can't find the solution, I'll be back.

[Jared Greathouse]

Maybe email Clement.

[Benoit Decoco]

My bad, for future people having the same problem, there is three auxiliary packages needed to perform the diff-and-diff : moremata, ftools, and reghdfe.