Remove implausible weight values using 7-step moderate EWMA flow.

Usage

remove_mod_ewma_wt(
  full_inc_df,
  exc_label = "Exclude-A-Traj-Moderate",
  max_rounds = 100,
  wtallow_formula = "piecewise",
  ewma_window = 15,
  mod_ewma_f = 0.75,
  perclimit_low = 0.5,
  perclimit_mid = 0.4,
  perclimit_high = 0,
  endpair_gap_days = 90
)

Arguments

full_inc_df: Data frame for one subject with columns: internal_id, age_days, ageyears, meas_m (weight in kg). Must have >= 3 rows.
exc_label: Character prefix for exclusion codes
max_rounds: Maximum number of rounds (default 100)
wtallow_formula: Formula for wtallow ("piecewise", "piecewise-lower", "allofus15")
ewma_window: Maximum number of positions on each side included in the EWMA (default 15)
mod_ewma_f: Directional factor: both directional (before/after) deviations must exceed mod_ewma_f x wtallow (default 0.75)
perclimit_low, perclimit_mid, perclimit_high: Percentage-limit settings passed to compute_perc_limit()
endpair_gap_days: Max within-pair age gap (days) for the end-pair median guard (default 90).

Value

Named character vector: internal_id -> exclusion code for excluded values.