Package: oalasso 1.0.0

Daijiro Kabata

oalasso: Outcome-Adaptive Lasso Propensity Scores

Estimates propensity scores by the outcome-adaptive lasso of Shortreed and Ertefaie (2017) <doi:10.1111/biom.12679> and the generalized outcome-adaptive lasso (GOAL) of Balde, Yang and Lefebvre (2023) <doi:10.1111/biom.13683>, using 'glmnet' with an exact penalty-scale correction so that the published objectives and tuning grids are reproduced. Tuning is by the weighted absolute mean difference balance criterion. The resulting score is designed to be supplied directly to the matchit() function of 'MatchIt' as a distance measure, to the weightit() function of 'WeightIt' as a propensity score, or to the psave() function of 'psAve' as an appended candidate.

Authors:Daijiro Kabata [aut, cre, cph]

oalasso_1.0.0.tar.gz
oalasso_1.0.0.zip(r-4.7-any)oalasso_1.0.0.zip(r-4.6-any)oalasso_1.0.0.zip(r-4.5-any)
oalasso_1.0.0.tgz(r-4.6-any)oalasso_1.0.0.tgz(r-4.5-any)
oalasso_1.0.0.tar.gz(r-4.7-any)oalasso_1.0.0.tar.gz(r-4.6-any)
oalasso_1.0.0.tgz(r-4.6-emscripten)
manual.pdf |manual.html✨
DESCRIPTION |NEWS
card.svg |card.png
oalasso/json (API)

# Install 'oalasso' in R:
install.packages('oalasso', repos = c('https://kabajiro.r-universe.dev', 'https://cloud.r-project.org'))

Bug tracker:https://github.com/kabajiro/oalasso/issues

Pkgdown/docs site:https://kabajiro.github.io

On CRAN:

Conda:

causal-inferenceepidemiologypropensity-scorevariable-selection

4.48 score 4 exports 30 dependencies

Last updated from:bc094196ca. Checks:9 OK. Indexed: yes.

TargetResultTimeFilesSyslog
linux-devel-x86_64OK254
source / vignettesOK274
linux-release-x86_64OK217
macos-release-arm64OK154
macos-oldrel-arm64OK153
windows-develOK172
windows-releaseOK184
windows-oldrelOK120
wasm-releaseOK112

Exports:oaloal_matchoal_wamdoal_weight

Dependencies:argclicobaltcodetoolscpp11farverforeachggplot2glmnetgluegridExtragtableisobanditeratorslabelinglatticelifecycleMatrixR6RColorBrewerRcppRcppEigenrlangS7scalesshapesurvivalvctrsviridisLitewithr

Getting Started with oalasso
Introduction: which covariates belong in a propensity score model? | Installation | A first look: instruments are excluded, confounders are kept | Reading the printout | Workflow 1: matching | Workflow 2: weighting | How the tuning works: the wAMD criterion | summary(), balance, and the other accessors | GOAL, in one paragraph | Practical policies worth knowing | Where to go next | References
Rendered fromoalasso.Rmdusingknitr::rmarkdown

Last update: 2026-07-02
Started: 2026-07-02

Method details and provenance
Notation | The OAL estimator | Step 1: the outcome (weight) model | Step 2: the adaptively penalized treatment model | Step 3: the tuning grid | Step 4: the wAMD selection criterion | Step 5: the propensity score | GOAL: the generalized outcome-adaptive lasso | Standardization protocol (fixed, not an argument) | The exact penalty-scale correction (why glmnet alone is not enough) | Provenance of every default | Differences from the author's legacy code | Relation to other software | A verified scale difference against the historical lqa-based references | Beyond linearity: honest notes on nonlinear extensions | The linear limitation | The outcome.coef hook | Experimental: importance-based weights (a recipe, not a feature) | References
Rendered frommethod-details.Rmdusingknitr::rmarkdown

Last update: 2026-07-02
Started: 2026-07-02

Using oalasso with psAve
Two packages, two kinds of propensity score model uncertainty | When this composition helps | A worked example | The criterion-interaction caveat | Validation status | References
Rendered frompsave-integration.Rmdusingknitr::rmarkdown

Last update: 2026-07-02
Started: 2026-07-02