mimrgns – problem of including ‘.class’ file format

[Wolf Edler]

I would like to calculate average marginal effects after running a linear regression with multiple imputed values.
I'm accessing my data by using a remote access. This just allows me to include do-files and ado-files. I found the mimrgns ado-file but realized when including the ado-file that there are 3 files to include (1) mimrgns.ado, (2) mimrgns_estimate.ado and (3) mimrgns_work.class
As I can't use the file format '.class' when using the remote access I wonder if I can skip the inclusion of 'mimrgns_work.class' and just include the mimrgns.ado and mimrgns_estimate.ado in my do-file program.
The command seems to work also without mimrgns_work.class. What does mimrgns_work.class do? When do I need 'mimrgns_work.class'?

My relevant commands are

adopath ++${prog}
which mimrgns.ado
which mimrgns_estimate.ado
*which mimrgns_work.class <-- I cannot include this when using the remote access

mi estimate, saving(miestfile, replace) esample(esample): xtreg logwage i.man##c.unemployment i.education , fe robust
mimrgns using miestfile, esample(esample) dydx(unemployment) at(unemployment=(0(90)720) man=(0 1))

[daniel klein]

mimrgns is (most likely) from SSC.

The command seems to work also without mimrgns_work.class.

I would be very surprised if it did. What makes you think so?

What does mimrgns_work.class do?

At the moment, pretty much everything. More specifically, it does all the syntax parsing, calls mi estimate (mimrgns_estimate.ado), combine the results (for pairwise comparisons), and reports the results. You can actually look at the code yourself

findfile mimrgns_work.class
doedit `"`r(fn)'"'

When do I need 'mimrgns_work.class'?

You do not need to call it, but mimrgns.ado does. And, mimrgns_estimate.ado expects an instance of that class to be passed as its only argument.

If there is really no possibility to get mimrgns_work.class installed on the remote machine, I believe your best option is to write a simple wrapper for you specific estimation command yourself. This could be as simple as

program mymargins , eclass properties(mi)
    version 15
    xtreg logwage i.man##c.unemployment i.education, fe robust
    margins , dydx(unemployment) at(unemployment=(0(90)720) man=(0 1)) post
end

Then call your command as

mi estimate : mymargins foo

(you may replace foo with any string of your choice)

This is what mimrgns does at its core. Of course, you will not have all the bells and whistles, e.g., you will get the standard mi estimate output, marginsplot might not work or produce wrong results etc., but for your situation, this is probably your best option.

Best
Daniel

[Wolf Edler]

Thank you very much for your fast response!