MRM-framework analyses on Toronto Police Service (TPS) open data

Four callables for TPS public-release crime-incident CSVs, used in the MRM empirical companion paper.

Value

Each mrm_tps_*() callable returns a named list with the computed statistic (Pareto exponent, Moran's I, or survival curve) and a plain-language interpretation; mrm_tps_load_hawkes_refit() returns the parsed Hawkes-refit manifest as a list.

Details

Functions:

• mrm_tps_levy_scaling(): Hill-MLE Pareto exponent of inter-incident step-length distribution on the lat/long-coded event stream.

• mrm_tps_moran_clustering(): global Moran's I + DBSCAN cluster summary on the lat/long-coded event stream.

• mrm_tps_neighbourhood_recurrence_km(): Kaplan-Meier inter-event gap distribution per HOOD_158 neighbourhood.

• mrm_tps_load_hawkes_refit(): convenience loader that pulls the precomputed per-category Hawkes (Markovian + Weibull/sin) fits from the paper_hawkes_refit.json manifest if available.

Examples

if (FALSE) {
  tps <- read.csv("Assault_Open_Data.csv")
  mrm_tps_levy_scaling(tps)
}