Dear Statalisters,
I use CEM with Parallel with the following command

Code:
parallel setclusters 4, force
parallel: cem first_obs(#0) age(#0)  average_weekly_sales (0 0.025 0.05 0.1 0.2 0.3 0.4 0.5 0.75 1 5 10 25 50), noimbal
To state verbally, I exactly match on the first observation and age and group the observation into stratas without specifying the treatment.
However, when I take a look at the stratas (produced as cem_strata by the Stata command) I see that there is a variation in first_obs and age although I exactly merge on first_obs and age.
Any other idea about how to solve the issue? Or alternatives to parallel command?