dml_core/lib.rs
1//! # dml-core
2//!
3//! Double / debiased machine learning (Chernozhukov et al., 2018) in Rust.
4//!
5//! ## Models
6//!
7//! - [`Plr`]: partially linear regression (partialling-out or IV-type score)
8//! - [`Irm`]: interactive regression model for a binary treatment (ATE / ATTE)
9//! - [`Pliv`]: partially linear IV regression
10//! - [`Iivm`]: interactive IV model, local average treatment effect (LATE) with a binary instrument
11//! - [`Did`] / [`DidCs`]: doubly robust difference-in-differences for panel data and
12//! repeated cross-sections (Sant'Anna & Zhao, 2020)
13//! - [`StaggeredDid`]: DID with staggered adoption, group-time effects `ATT(g, t)` and their
14//! aggregation by cohort, calendar time or event time (Callaway & Sant'Anna, 2021)
15//!
16//! All of them have scores that are linear in θ and share one engine
17//! ([`fit_model`]): repeated cross-fitting, DML2 estimation, standard errors,
18//! confidence intervals, a multiplier [`Bootstrap`] for joint inference and
19//! multiple-testing adjustment. New scores plug in via [`LinearScoreModel`].
20//!
21//! ## Heterogeneous effects
22//!
23//! - [`SLearner`], [`TLearner`], [`XLearner`], [`RLearner`], [`DrLearner`]: CATE meta-learners
24//! - [`Blp`]: best linear predictor of the CATE and group average effects (GATE) with
25//! robust inference, from [`DrLearner`] fits or [`DmlFit::cate`] / [`DmlFit::gate`]
26//!
27//! ## Building blocks
28//!
29//! - [`DmlData`]: outcome `y`, one or more treatments `d`, covariates `x`, optional instrument `z`
30//! - [`Regressor`] / [`Classifier`]: plug-in nuisance learners
31//! ([`Ols`], [`Lasso`], [`LogisticRegression`], [`RandomForestRegressor`],
32//! [`RandomForestClassifier`], [`GradientBoostingRegressor`], [`GradientBoostingClassifier`]
33//! built in)
34//! - [`KFold`] / [`SampleSplitting`]: reproducible sample splitting
35//!
36//! ```
37//! use dml_core::{DmlData, Ols, Plr};
38//! use ndarray::{Array1, Array2};
39//!
40//! let n = 200;
41//! let x = Array2::from_shape_fn((n, 2), |(i, j)| ((i * 7 + j * 3) % 11) as f64);
42//! let d: Array1<f64> = (0..n).map(|i| x[[i, 0]] + ((i * 13) % 5) as f64).collect();
43//! let y: Array1<f64> = (0..n).map(|i| 0.5 * d[i] - x[[i, 1]] + ((i * 17) % 3) as f64).collect();
44//!
45//! let fit = Plr::new(Ols::new(), Ols::new()).n_folds(5).seed(42).fit(&DmlData::new(y, d, x)?)?;
46//! assert!((fit.coef()[0] - 0.5).abs() < 0.1);
47//! let ci = fit.confint(0.95)?; // shape (n_treat, 2)
48//! # let _ = ci;
49//! # Ok::<(), dml_core::DmlError>(())
50//! ```
51
52pub mod blp;
53pub mod bootstrap;
54pub mod cate;
55pub mod crossfit;
56pub mod data;
57pub mod error;
58pub mod fit;
59pub mod learner;
60pub(crate) mod linalg;
61pub mod model;
62pub mod score;
63pub mod staggered;
64pub mod stats;
65
66pub use blp::Blp;
67pub use bootstrap::{Bootstrap, BootstrapMethod};
68pub use cate::{CateFit, DrLearner, RLearner, SLearner, TLearner, XLearner};
69pub use crossfit::{
70 cross_fit_predict, cross_fit_predict_proba, cross_fit_predict_proba_subset,
71 cross_fit_predict_subset, Fold, KFold, SampleSplitting,
72};
73pub use data::DmlData;
74pub use error::{DmlError, Result};
75pub use fit::{fit_model, DmlFit, PAdjust};
76pub use learner::{
77 Classifier, GradientBoostingClassifier, GradientBoostingRegressor, Lasso, LogisticRegression,
78 MaxFeatures, Ols, RandomForestClassifier, RandomForestRegressor, Regressor,
79};
80pub use model::{
81 Did, DidCs, DidScore, Iivm, Irm, IrmScore, LinearScoreModel, Pliv, Plr, PlrScore,
82 ScoreEstimate, ScoreInput,
83};
84pub use score::LinearScore;
85pub use staggered::{
86 Aggregation, ControlGroup, DidAggregation, GroupTime, GtCombinations, PanelData,
87 RepeatedCrossSections, StaggeredDid, StaggeredDidFit, StaggeredInput,
88};