Error with coefplot about fixest

[phisherblack]

Awesome! thank you some much! i'll get the Dev version.

[lrberge]

Hi, I have a small idea, but I need confirmation. Could you send me data/code that leads to this error?

[phisherblack]

Thanks,
here is the code i am trying to run that is throwing the error.

mod1.full<-feols(log.gdp ~ success::year(1981)+log.gdppc+log.pop+bmr_dem+ armed_conflict| ccode + year, data = kcoup)

coefplot(mod1.full).

is this a bug with the package?

[lrberge]

Aren't there coefficients that are removed from the estimation because of collinearity?

Could you make me a link to (or dm me) the data, so that I can run it on my computer and replicate the problem?

[phisherblack]

Thank you. I have DM'ed you with the dataset.

…

On Mon, Sep 28, 2020, 8:07 AM Laurent Bergé ***@***.***> wrote: Aren't there coefficients that are removed from the estimation because of collinearity? Could you make me a link to (or dm me) the data, so that I can run it on my computer and replicate the problem? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#45 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACZD44KSRWC6ZX33YARMVN3SIB4ADANCNFSM4R3ZPS7A> .

[lrberge]

Thanks, that was a bug when interacted coefficients were removed from the estimation due to NAness. Now fixed.

The dev version should work.