Struct PolicyTree 

pub struct PolicyTree { /* private fields */ }

A fitted policy tree.

Implementations

impl PolicyTree

pub fn fit( signal: ArrayView1<'_, f64>, features: ArrayView2<'_, f64>, depth: usize, min_samples_leaf: usize, ccp_alpha: f64, ) -> Result<Self>

Fit on the orthogonal signal (one value per unit) and the features the policy may depend on. DoubleML’s defaults: min_samples_leaf = 8, ccp_alpha = 0.01.

pub fn depth(&self) -> usize

pub fn min_samples_leaf(&self) -> usize

pub fn ccp_alpha(&self) -> f64

pub fn n_leaves(&self) -> usize

Number of leaves (distinct recommendations groups).

pub fn predict(&self, features: ArrayView2<'_, f64>) -> Result<Array1<f64>>

Recommended treatment (1) or not (0) for each row of features: the class with the larger weighted share in the row’s leaf (no treatment on ties, as scikit-learn’s argmax).

pub fn nodes(&self) -> Vec<PolicyNode>

The nodes in preorder; node 0 is the root (for plotting the tree).

pub fn export_text(&self, feature_names: Option<&[String]>) -> String

The decision rules as text, in the style of scikit-learn’s export_text; feature_names defaults to x0, x1, ….

Trait Implementations

impl Clone for PolicyTree

fn clone(&self) -> Self

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for PolicyTree

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.