API reference

Every function, by family. Each has its own page: signature, parameters, returns, notes, references and examples.

Disparities

Veil of darkness, odds ratios and risks, the outcome test, benchmarks and dyadic rates.

Function	What it computes
`vodtst()`	Veil-of-darkness test
`orrisk()`	Risks behind an odds ratio
`outtst()`	Outcome test
`dspbm()`	Benchmark decomposition of a disparity
`dyadrt()`	Dyadic offending rates

Forensics

DNA match statistics, the weight and calibration of likelihood ratios, kinship, paternity, mixtures and database searches.

Function	What it computes
`gtmp()`	Genotype match probability
`lrpost()`	Posterior probability from a likelihood ratio
`cllr()`	Log-likelihood-ratio cost
`cllrmin()`	Log-likelihood-ratio cost after recalibration
`lrcal()`	Calibrated likelihood ratios
`tipplt()`	Tippett plot
`eceplt()`	Empirical cross-entropy plot
`mixlr()`	Mixture likelihood ratio
`kinidx()`	Kinship index
`patidx()`	Paternity index
`dbsrch()`	Database search

Policing

Clearance rates, exposure proxies, misrecorded groups and regression to the mean.

Function	What it computes
`clrsmp()`	Clearance-rate decomposition and Simpson’s paradox
`expbnd()`	Rates and disparities when exposure is measured by a proxy
`misrd()`	Risk difference when group membership is misrecorded
`rtmhs()`	Regression to the mean at selected hot spots

Corrections

Incapacitation, release rules, time in custody, sentence effects and hazards after selection.

Function	What it computes
`incap()`	Incapacitation: crime rate, share prevented and the marginal year
`incapr()`	Incapacitation under desistance and replacement
`sellab()`	Evaluating a release rule from the released alone
`losbnd()`	Mean time in custody from completed and current stays
`sntbnd()`	Bounds on a sentence’s effect without assumptions
`hzdsel()`	Period-two hazards after selective depletion

Networks

Co-offending graphs, centrality, cores, components, communities, brokerage and the Cheeger bounds.

Function	What it computes
`coofg()`	Co-offending graph from an incident table
`ntdeg()`	Degree and strength of each node
`ntbtw()`	Betweenness of each node
`ntcls()`	Closeness of each node
`ntevc()`	Eigenvector centrality of each node
`ntcore()`	k-core of each node
`ntcmp()`	Connected components
`ntlvn()`	Communities by the Louvain method
`ntmod()`	Modularity of a partition
`ntbrk()`	Brokerage roles
`ntchg()`	Cheeger bounds on a network’s bottleneck

Spatial

Distances and projection, centrography, nearest neighbours and K functions, hot spots, space-time tests and journey to crime.

Function	What it computes
`hvdist()`	Great-circle distance
`eqproj()`	Equal-area projection of longitude and latitude, and its inverse
`cgstat()`	Mean centre and dispersion of points
`cgmedn()`	Median centre of points
`webpt()`	Centre of minimum distance
`sdell()`	Standard deviational ellipse
`cvhull()`	Convex hull of points
`nnidx()`	Nearest-neighbour index
`ripk()`	Ripley’s K function
`ripenv()`	Envelopes of Ripley’s L function and a test of complete spatial randomness
`fzmode()`	Fuzzy mode
`kmpart()`	K-means partition of points
`nnhclu()`	Nearest-neighbour hierarchical clustering
`stacel()`	Search-circle hot-spot areas
`rnnhcl()`	Risk-adjusted nearest-neighbour hierarchical clustering
`hdbang()`	Head-bang smoothing
`crwalk()`	Correlated random walk of a series of events
`mantst()`	Mantel’s space-time test
`stmavg()`	Spatial-temporal moving average
`jtdecay()`	Distance-decay functions for journey to crime
`jtsurf()`	Journey-to-crime surface
`jtcal()`	Empirical distance decay from solved trips
`jtacc()`	Accuracy of a journey-to-crime surface on a solved series
`bjtc()`	Bayesian journey to crime
`spwt()`	Spatial weights
`gearc()`	Geary’s C
`gorg()`	The general G of Getis and Ord
`lmorn()`	Local Moran (LISA)
`locgi()`	Local G and G* of Getis and Ord
`dirmn()`	Directional mean and circular variance

Dossier methods

Fairness of risk tools, trace evidence, aoristic time, Benford screens, glass grouping and transfer, constant reporters, sensitivity bounds, trajectory models, lowess, tool marks, grid Gi* and the point pattern test.

Function	What it computes
`fair()`	Fairness metrics of a binary risk tool by group
`tlvc()`	Two-level model for multivariate trace evidence
`aorwt()`	Aoristic weights
`bdig()`	Benford digit screens
`skp()`	Scott-Knott grouping of measurements
`rptc()`	Constant reporters of a reporting panel
`rswp()`	Rosenbaum sensitivity bounds for matched pairs
`zgll()`	Group-based trajectory model for counts
`mvtll()`	Group-based multivariate trajectories
`tfrp()`	Glass fragments recovered under transfer and persistence
`lwss()`	Lowess smoothing
`chmb()`	The Chumbley score for two tool-mark profiles
`gstr()`	Local Getis-Ord Gi* on a grid of counts
`sptst()`	Andresen’s spatial point pattern test

Checks

Which labelling of a table reproduces a reported contrast, and an audit of categorical inputs.

Function	What it computes
`relchk()`	Which labelling reproduces a reported contrast?
`catchk()`	Audit categorical inputs before an analysis

Counts, rates and tables

Rate ratios from count regression with exposure and clusters; rates and proportions by group with exact and approximate intervals; direct and indirect standardisation; cross-tabulation with tests and measures of association.

Function	What it computes
`cntreg()`	Count regression: rate ratios between groups
`grprr()`	Rates and proportions by group: ratios and differences
`dastd()`	Directly standardised rates
`sirex()`	Indirect standardisation: standardised ratios with shrinkage
`ctabx()`	Cross-tabulation with tests and measures of association

Presets

A published dataset’s columns onto a generic function’s arguments.

Function	What it computes
`clprep()`	Presets: a published dataset’s columns onto a generic function’s arguments

Learning

The elastic net with cross-validation, logistic regression and random forests; the learners of the cross-fitted estimators.

Function	What it computes
`enpath()`	Elastic net by coordinate descent, with cross-validation
`lgtreg()`	Logistic regression
`rfor()`	Random forest for regression and probabilities

Causal estimates

Double machine learning by cross-fitting, with the learners above; augmented inverse probability weighting and targeted maximum likelihood from the same nuisances.

Function	What it computes
`dmlx()`	Double machine learning by cross-fitting

Proofs

Function	What it computes
`proof()`	Machine-checked proofs behind the package

Package

Function	What it computes
`crimlab-package()`	crimlab: Criminology and Sociolegal Statistics