Add support for sample_weights in DRTester#1005

Open

gdaiha wants to merge 27 commits into

py-why:mainfrom

gdaiha:gdaiha/add-sample-weights-drtester-signed

Commits on Sep 11, 2025

created weighted utils module inside econml.validate
gdaiha
committed
fixed weighted_stat module and standard error calculations in dr_tester
gdaiha
committed
fixed standard error calculations in uplift
gdaiha
committed
fixed n_size_bootstrap in bootstraping toc and qini curves
gdaiha
committed
adjusted weighted averages in drtester
gdaiha
committed
removed weighted se in weighted_utils
gdaiha
committed
adjusted standard error calculation in calc_uplift
gdaiha
committed
readded weighted_se and applied in drtester and calc_uplift
gdaiha
committed
Merge branch 'main' into gdaiha/add-sample-weights-drtester-signed
gdaiha
authored
standardized drtester vals_ to two dimensions
gdaiha
committed
Merge branch 'gdaiha/add-sample-weights-drtester-signed' of https://github.com/gdaiha/EconML into gdaiha/add-sample-weights-drtester-signed
gdaiha
committed

Commits on Sep 17, 2025

fixed treatments in drtester to allow controls different than zero
gdaiha
committed

Commits on Sep 19, 2025

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Add support for sample_weights in DRTester#1005

Add support for sample_weights in DRTester#1005
gdaiha wants to merge 27 commits into
py-why:mainfrom
gdaiha:gdaiha/add-sample-weights-drtester-signed

Commits on Sep 10, 2025

added support for sample_weight in drtester

changed numpy typing in econml results

fixed sample_weight pass in fit_nuisance_train on validation sample

implemented weighted quantiles and percentiles to avoid dependency on numpy>1.22

Commits on Sep 11, 2025

created weighted utils module inside econml.validate

fixed weighted_stat module and standard error calculations in dr_tester

fixed standard error calculations in uplift

fixed n_size_bootstrap in bootstraping toc and qini curves

adjusted weighted averages in drtester

removed weighted se in weighted_utils

adjusted standard error calculation in calc_uplift

readded weighted_se and applied in drtester and calc_uplift

Merge branch 'main' into gdaiha/add-sample-weights-drtester-signed

standardized drtester vals_ to two dimensions

Merge branch 'gdaiha/add-sample-weights-drtester-signed' of https://github.com/gdaiha/EconML into gdaiha/add-sample-weights-drtester-signed

Commits on Sep 17, 2025

fixed treatments in drtester to allow controls different than zero

Commits on Sep 19, 2025

fix assign units to validation sets in evaluate_cal to avoid same value in two groups

changed pvalues to be two-sided in evaluate_uplift

added cross_val_predict_with_weights to drtester

guarantee support for cross_val_predict_with_weights to sklearn>=1.0

written tests for sample_weights support in drtester

removed cast sample_weight to integer and changed to a normalization by minimum value

Commits on Sep 22, 2025

cast weights vector to float to avoid typing error in numpy

cast weights vector to float to avoid typing error in numpy

added comment in treatment loop

Commits on Dec 18, 2025

Merge branch 'main' into gdaiha/add-sample-weights-drtester-signed

Commits on Apr 10, 2026

Merge branch 'main' into gdaiha/add-sample-weights-drtester-signed