Test for parallell trends

[Torje Os]

Hello!
After using xdidregress we are trying to use estat ptrends to test the trends (some of the plots are a bit sketchy, and, thus, we wanted to do a test as well). The problem is that we get an error message that states: treatment assignment vary; not allowed with estat ptrend. Is there a general way to "correct" this error, or is there another way to do it? Unfortunately we cannot display our data as it is confidential.
Thank you in advance.

[Jared Greathouse]

I don't care if your data are confidential, you still have to (as per the FAQ!!!) present a minimal worked example. Simulate a dataset that looks just like yours, but with fake numbers. Once you can present that, we can get to helping you easier.

It sounds like Stata is unhappy with the treatment coming and going- but I don't know, I need to see how your data look first. So, falsify your dataset first and then

dataex , count(300)

and let's see what we're looking at