Mirrors the canonical aggregation schemes (overall ATT, by-cohort,
by-calendar-time, by-event-time) and produces a tidy
data.frame consumed by the rmorie / MRM downstream
pipelines.
Usage
morie_did_aggregate_gt_att(
gt_results,
aggregation = "overall",
time_col = "time",
cohort_col = "cohort",
att_col = "att",
se_col = "std_error"
)Arguments
- gt_results
Output of
morie_did_group_time_att.- aggregation
One of
"overall"(default),"cohort","calendar_time","event_time".- time_col, cohort_col, att_col, se_col
Column-name overrides.
Examples
gt <- data.frame(cohort = c(2, 2, 3, 3),
time = c(2, 3, 3, 4),
att = c(1.0, 1.5, 2.0, 2.5),
std_error = c(0.4, 0.4, 0.3, 0.3))
out <- morie_did_aggregate_gt_att(gt, aggregation = "overall")
out$estimate
#> [1] 1.75
