Learn R Programming

vimp (version 2.3.3)

Perform Inference on Algorithm-Agnostic Variable Importance

Description

Calculate point estimates of and valid confidence intervals for nonparametric, algorithm-agnostic variable importance measures in high and low dimensions, using flexible estimators of the underlying regression functions. For more information about the methods, please see Williamson et al. (Biometrics, 2020), Williamson et al. (JASA, 2021), and Williamson and Feng (ICML, 2020).

Copy Link

Version

Install

install.packages('vimp')

Monthly Downloads

318

Version

2.3.3

License

MIT + file LICENSE

Issues

Pull Requests

Stars

Forks

Repository

https://github.com/bdwilliamson/vimp

Homepage

https://bdwilliamson.github.io/vimp/

Maintainer

Brian D. Williamson

Last Published

August 28th, 2023

Functions in vimp (2.3.3)

estimate_eif_projection

Estimate projection of EIF on fully-observed variables

extract_sampled_split_predictions

Extract sampled-split predictions from a CV.SuperLearner object

estimate_type_predictiveness

Estimate Predictiveness Given a Type

vimp: Perform Inference on Algorithm-Agnostic Intrinsic Variable Importance

Return an estimator on a different scale

Shapley Population Variable Importance Measure (SPVIM) Estimates and Inference

Nonparametric Intrinsic Variable Importance Estimates and Inference

Return test-set only data

Create Folds for Cross-Fitting

Estimate ANOVA decomposition-based variable importance.

estimate_nuisances

Estimate nuisance functions for average value-based VIMs

Estimate area under the receiver operating characteristic curve (AUC)

process_arg_lst

Process argument list for Super Learner estimation of the EIF

Print vim objects

measure_deviance

Estimate the deviance

Estimate mean squared error

Nonparametric Intrinsic Variable Importance Estimates: Deviance

vimp_hypothesis_test

Perform a hypothesis test against the null hypothesis of \(\delta\) importance

predictiveness_measure

Construct a Predictiveness Measure

print.predictiveness_measure

Print predictiveness_measure objects

Run a Super Learner for the provided subset of features

Nonparametric Intrinsic Variable Importance Estimates: Classification accuracy

measure_accuracy

Estimate the classification accuracy

Turn folds from 2K-fold cross-fitting into individual K-fold folds

Create necessary objects for SPVIMs

get_cv_sl_folds

Get a numeric vector with cross-validation fold IDs from CV.SuperLearner

Obtain the type of VIM to estimate using partial matching

measure_average_value

Estimate the average value under the optimal treatment rule

measure_cross_entropy

Estimate the cross-entropy

Influence function estimates for SPVIMs

Estimate variable importance standard errors

Nonparametric Intrinsic Variable Importance Estimates: ANOVA

Neutralization sensitivity of HIV viruses to antibody VRC01

Standard error estimate for SPVIM values

Nonparametric Intrinsic Variable Importance Estimates: AUC

Confidence intervals for variable importance

format.predictiveness_measure

Format a predictiveness_measure object

Format a vim object

Nonparametric Intrinsic Variable Importance Estimates and Inference using Cross-fitting

measure_r_squared

Estimate R-squared

vimp_regression

Nonparametric Intrinsic Variable Importance Estimates: ANOVA

Nonparametric Intrinsic Variable Importance Estimates: R-squared

Merge multiple vim objects into one

estimate.predictiveness_measure

Obtain a Point Estimate and Efficient Influence Function Estimate for a Given Predictiveness Measure

est_predictiveness_cv

Estimate a nonparametric predictiveness functional using cross-fitting

Estimate a Predictiveness Measure

Create complete-case outcome, weights, and Z

check_fitted_values

Check pre-computed fitted values for call to vim, cv_vim, or sp_vim

Compute bootstrap-based standard error estimates for variable importance

Average multiple independent importance estimates

Check inputs to a call to vim, cv_vim, or sp_vim

est_predictiveness

Estimate a nonparametric predictiveness functional