Crate oaxaca_blinder 

A Rust implementation of the Oaxaca-Blinder decomposition method.

This library provides tools to decompose the mean difference in an outcome variable between two groups into an “explained” part (due to differences in observable characteristics) and an “unexplained” part (due to differences in the returns to those characteristics).

Currently, the library supports numerical predictors and calculates standard errors using bootstrapping.

Example

use polars::prelude::*;
use oaxaca_blinder::OaxacaBuilder;

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let df = df!(
        "wage" => &[10.0, 12.0, 11.0, 13.0, 15.0, 20.0, 22.0, 21.0, 23.0, 25.0],
        "education" => &[12.0, 16.0, 14.0, 16.0, 18.0, 12.0, 16.0, 14.0, 16.0, 18.0],
        "gender" => &["F", "F", "F", "F", "F", "M", "M", "M", "M", "M"]
    )?;

    let results = OaxacaBuilder::new(df, "wage", "gender", "F")
        .predictors(&["education"])
        .run()?;

    results.summary();
    Ok(())
}

Quantile Regression Decomposition

use polars::prelude::*;
use oaxaca_blinder::QuantileDecompositionBuilder;

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let df = df!(
        "wage" => &[10.0, 12.0, 11.0, 13.0, 15.0, 20.0, 22.0, 21.0, 23.0, 25.0, 9.0, 18.0],
        "education" => &[12.0, 16.0, 14.0, 16.0, 18.0, 12.0, 16.0, 14.0, 16.0, 18.0, 10.0, 20.0],
        "gender" => &["F", "F", "F", "F", "F", "F", "M", "M", "M", "M", "M", "M"]
    )?;

    let results = QuantileDecompositionBuilder::new(df, "wage", "gender", "F")
        .predictors(&["education"])
        .quantiles(&[0.25, 0.5, 0.75])
        .run()?;

    results.summary();
    Ok(())
}

Re-exports

pub use crate::quantile_decomposition::QuantileDecompositionBuilder;

pub use crate::jmp::decompose_changes;

pub use crate::dfl::run_dfl;

pub use crate::heckman::heckman_two_step;

pub use crate::akm::AkmBuilder;

pub use crate::akm::AkmResult;

pub use crate::matching::engine::MatchingEngine;

Modules

akm

dfl

formula

heckman

jmp

matching

quantile_decomposition

Machado-Mata Quantile Regression Decomposition

Structs

BudgetAdjustment

Represents a recommended adjustment for an individual to improve pay equity.

ComponentResult

Represents the calculated result for a single component or variable.

DecompositionDetail

Represents a component of the decomposition (e.g., two-fold or three-fold).

OaxacaBuilder

The main entry point for configuring and running an Oaxaca-Blinder decomposition.

OaxacaResults

Holds all the results from the Oaxaca-Blinder decomposition.

TwoFoldResults

Holds results for the two-fold decomposition, including detailed components.

Enums

OaxacaError

Error type for the oaxaca_blinder library.

ReferenceCoefficients

Represents the choice of reference coefficients for the two-fold decomposition.