eventstudy2 : downward compatibility issues to Stata 14 + solutions

[Thomas Kaspereit]

Issue:

I am stuck at the very first stage of using your package: Test running your sample code with your sample data.

The very first sample code yields “invalid syntax,” without any further explanation. In fact, I found a questions exactly same as mine on Statalist dated October 2022, and you replied to him to post a screenshot after setting trace on, but the original poster did not post any further questions.

So, I set trace on, and send you the screenshot of what Stata yields. I didn’t find any clue to fix my issue, and I was hoping if you could see what’s going on?


Solution:

The problem was the number of options. I counted the number of options in the eventstudy2 definition, and there were 70 options.

So I erased options saveram and crossonly [Thomas Kaspereit: in the eventstudy2.ado file]. I am not sure what they do but I thought they are irrelevant for analysis itself [Thomas Kaspereit: they are not crucial].

When I run the code after deleting those two options, then it goes beyond the initial error message. So “invalid syntax” should have been from the number of options being 70.

But I still had more errors.

It said nearmrgstable is already defined. So I cut out the part that defines nearmrgstable and saved it as a new ado file.

There were further error messages asking me to install user defined programs such as moremata, distinct, _gprod, rmse, and parallel.

After installing all user written programs as asked by Stata, the code finally ran.

I haven’t checked the output yet, but am happy to see it running after all.