Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/dmolitor/strike

Rust implementation of 1:1, nearest neighbor, propensity score matching.
https://github.com/dmolitor/strike

Last synced: 18 days ago
JSON representation

Rust implementation of 1:1, nearest neighbor, propensity score matching.

Awesome Lists containing this project

README 

        
# strike



Strike implements 1:1, nearest neighbor, propensity score matching with replacement.

The target estimand is the Average Treatment Effect on the Treated (ATT),

which is the average effect of treatment for those who receive treatment:

$$\frac{1}{N^T}\sum_{i = 1}^{N^T}(Y_i(1) - Y_i(0)).$$ Strike also implements the

consistent estimator for the variance of the matching estimator for the ATT, as

proposed in "Large Sample Properties of Matching Estimators for Average

Treatment Effects" (Abadie and Imbens, 2006).



## Example



We can use propensity score matching (PSM) to estimate the effect of smoking on

psychological health using observational data. In this setup, the treatment

variable is smoking; the outcome variable is psychological distress, measured

on a unit interval ranging from 10 to 50. PSM matches (with replacement) each

smoker in the dataset to a non-smoker based on their propensity to be a smoker.

Propensity scores are estimated using logistic regression with a number of

potentially counfounding covariates, including sex, indigenous status,

education level (high school completion), marital status,

region of residence (major cities, inner regional or outer regional),

language background (English speaking or not), age, and risky alcohol

use (Yes/ No).



Strike will take care of all the implementation details under the hood as long

as all the covariates/treatment/outcome are numeric. To estimate the ATT,

we simply need to execute the Strike binary and pass it the path to the

data file as well as the names of the treatment indicator and outcome variable.



```rust

cargo build

 

cargo run -- ./examples/smoking.csv smoker psyc_distress

// STRIKE =======================================

// 

// # Treat: 974 | # Control (distinct): 736

//

// ATT                     : 3.386

// Variance                : 0.079

// 95% Confidence Interval : (2.835, 3.937)

```



On average, the effect of smoking among smokers is an increase in psychological

distress by ~3.4 units, with a 95% confidence interval that excludes 0.