Interactive Threshold Selection Based on Expected Number of Extrema

Helps determine appropriate hop_idx threshold by showing how many extrema would be retained at different threshold levels. Particularly useful when you have domain knowledge about the expected number of true extrema.

Usage

suggest.threshold(
  extrema_df,
  expected_n_extrema = NULL,
  expected_range = NULL,
  plot = TRUE,
  exclude_global = TRUE
)

Arguments

extrema_df: Data frame from create.hop.nbhd.extrema.df with hop_idx column
expected_n_extrema: Expected number of true extrema (NULL for no target)
expected_range: Optional vector c(min, max) for expected range
plot: Logical; create diagnostic plots (default TRUE)
exclude_global: Logical; exclude global extrema (hop_idx = Inf) from counts

Value

List with suggested thresholds and diagnostic information

Examples

extrema_df <- data.frame(
  vertex = 1:6,
  hop_idx = c(1, 2, 3, 4, Inf, 6),
  is_max = c(0, 1, 0, 1, 1, 0),
  label = c("m1", "M1", "m2", "M2", "M3", "m3")
)
suggestions <- suggest.threshold(
  extrema_df,
  expected_range = c(2, 3),
  plot = FALSE
)
suggestions$suggestions$recommended_threshold