Compute DiD and validate parallel trends

You are given panel-style records through three parallel arrays: period, group, and outcome. For each observation i, period[i] is a time label, group[i] is 1 for treated units and 0 for control units, and outcome[i] is the measured value. First, compute a difference-in-differences estimate. Let the selected pre-intervention time be: $$p_{pre} = \max\{\,period[i] \mid period[i] < 0\,\}$$ and let the selected post-intervention time be: $$p_{post} = \min\{\,period[i] \mid…