csdid error 3598

[Henry Strawforrd]

I am using the incredible command csdid by @avila Fernando Rios-Avila https://friosavila.github.io/playing...ain_csdid.html

I got the following error from one regression

matrix __00000E not found
stata():         3598 Stata returned error
makerif2():   - function returned error
<istmt>:       - function returned error

I got this after changing only the outcome relative to a previous regression. I checked the outcome and it has little variation, only 0.23% of observations in the sample are 1, the rest zero.

Might this be the reason of the error, that some groups have no variation in the outcome? Or what is the matrix __00000E?

Thanks a lot!

[FernandoRios]

hi Henry
that might be it. But would have to check in detail with some data example.
However, because that is the main change you do to the model, chances are that the model is not defined because of lack of outcome variation.
HTH
F

[Henry Strawforrd]

Okay, thank you