Skip to contents

Classify whether a spesim run matches the intended qualitative regime

Source: R/regime.R
spesim_regime.Rd

This helper turns the numeric outputs of spesim_audit() into a compact, teaching-friendly judgement (green/amber/red) about whether the realised simulation matches the qualitative regime the user believes they are simulating.

The defaults are deliberately conservative: when sample sizes are too small or signals are weak, the result tends toward amber rather than over- confident green/red.

Usage

spesim_regime(res, audit = NULL, thresholds = list())

Arguments

res

A spesim_result list (as returned by spesim_run()).

audit

Optional list returned by spesim_audit(). If NULL, the function will compute an audit internally.

thresholds

Optional named list overriding default thresholds. See Details.

Value

A named list with sub-lists spatial, filtering, sampling. Each contains:

  • grade: one of "green", "amber", "red", "unknown"

  • reason: short explanation

  • metrics: compact numeric context

Details

Default thresholds (you can override via thresholds=):

  • Spatial NN ratio vs CSR (per species; uses dominant A if present):

    • clustered: nn_ratio < 0.85

    • CSR_like: 0.85 ≤ nn_ratio ≤ 1.15

    • inhibited: nn_ratio > 1.15

    • minimum points for classification: min_n = 50

  • Filtering (per gradient-responsive species):

    • minimum occupied sites: min_occ = 5

    • good: Spearman rho(-|z|, abundance) > 0.2 AND slope(abund~z) < 0

    • opposite: rho < -0.2 OR slope > 0

  • Sampling constraint:

    • if boundary exclusion estimate > 50% → amber/red (scheme-dependent)

    • if random placement attempts are close to the cap (≥90% of max attempts) → amber