Build threshold guidance summaries for behavior classification — grz_behavior_threshold

Produces overall and hourly quantile summaries (plus diagnostic plots) for selected metrics to help users choose rule-based behavior thresholds.

Usage

grz_behavior_threshold_guide(
  data,
  metrics = c("step_m", "turn_rad"),
  cohort_col = NULL,
  tz_local = "UTC",
  probs = c(0.1, 0.25, 0.5, 0.75, 0.9),
  max_points_plot = 150000L,
  seed = 1,
  include_tuning = TRUE,
  tuning_rest_step_grid = seq(3, 12, by = 1),
  tuning_rest_speed_grid = seq(0.03, 0.09, by = 0.02),
  tuning_max_rows = 25000L,
  return_class = c("data.frame", "data.table")
)

Arguments

data: Input data containing at least sensor_id, datetime, lon, lat.
metrics: Metrics used for threshold guidance.
cohort_col: Optional cohort column used in summaries and facets.
tz_local: Time zone used to derive hour-of-day.
probs: Quantile probabilities used in summary tables.
max_points_plot: Maximum points used in density plot subsample.
seed: Random seed used for plotting subsample.
include_tuning: Logical; run threshold tuning diagnostics.
tuning_rest_step_grid: Candidate rest_step_max values for tuning.
tuning_rest_speed_grid: Candidate rest_speed_max values for tuning.
tuning_max_rows: Maximum rows used in tuning sweep.
return_class: Output class for returned tables.

Value

A list with overall, hourly, plots (ggplot objects), and optional tuning diagnostics.