Rust API

The Rust crate dml-core is documented on docs.rs/dml-core.

Main entry points:

Item	Purpose
DmlData	Outcome, treatment(s), covariates, optional instrument, time indicator and clusters
Plr, Lplr, Plpr, Irm, Pliv, Iivm, Did, DidCs, Ssm	Models (builders with n_folds, n_rep, seed, fit, fit_with_folds)
Apo, Apos, Pq, Lpq, Cvar, Qte	Average potential outcomes and quantile models
RdFlex, rd	Regression discontinuity with covariate adjustment, and the rdrobust back end
StaggeredDid	Staggered difference-in-differences
datasets	Simulated data sets (DoubleML's data generating processes)
DmlFit	Results: estimates, inference, bootstrap, sensitivity analysis, learner losses
Regressor, Classifier	Learner traits, implemented by the built-in learners
LinearScoreModel	Trait for adding a new orthogonal score

The crate builds with Rust 1.80 or newer. Parallel cross-fitting is enabled by the default parallel feature (rayon).