Impute a leading ovulation anchor before a participant's first recorded menses onset

Days observed BEFORE a participant's first recorded menses onset (the left-censored start of participation) have no cycle anchor on their left, so PACTS cannot place them: the luteal phase they belong to has a closing menses (the first onset) but no ovulation. This opt-in rule imputes that missing ovulation with the same population-average backward count the package already uses inside observed cycles: ovulation is placed luteal_days days (default 15) before the first menses onset, so the leading days scale as the end of a luteal phase in cyclic_time_impute / cyclic_time_imp_ov (never in the confirmed-only columns). The anchor is marked in a new column ovtoday_leading_impute (1 on the imputed day, 0 elsewhere). If the imputed day lies before the first observed row, a blank row is added on that date (other columns NA), exactly as impute_next_menses_onsets() adds a blank row for an imputed onset.

Usage

impute_leading_ovulation_anchors(
  data,
  id,
  date,
  menses,
  ovtoday,
  luteal_days = 15
)

Arguments

data, id, date, menses, ovtoday

As in pacts_scaling().

luteal_days

Integer; days before the first menses onset at which the ovulation is placed. Default 15 (the package's backward-count convention).

Value

data with an added integer column ovtoday_leading_impute, and possibly one added row per participant on the imputed date.

Details

Nothing is imputed when the participant has no menses onset, when no observed row precedes the first onset, or when a confirmed ovulation (ovtoday == 1) already lies before the first onset. Only the general rule is applied; study-specific gating is the caller's responsibility.