Lrnr_screener_importance: Variable Importance Screener
In jeremyrcoyle/sl3: Pipelines for Machine Learning and Super Learning

Description Format Value Parameters See Also Examples

This learner screens covariates based on their variable importance, where the importance values are obtained from the learner. Any learner with an importance method can be used. The set of learners with support for importance can be found with sl3_list_learners("importance"). Like all other screeners, this learner is intended for use in a Pipeline, so the output from this learner (i.e., the selected covariates) can be used as input for the next learner in the pipeline.

An R6Class object inheriting from Lrnr_base.

A learner object inheriting from Lrnr_base with methods for training and prediction. For a full list of learner functionality, see the complete documentation of Lrnr_base.

learner: An instantiated learner that supports variable importance. The set of learners with this support can be obtained via sl3_list_learners("importance").
num_screen = 5: The top n number of "most impotant" variables to retain.
...: Other parameters passed to the learner's importance function.

Other Learners: Custom_chain, Lrnr_HarmonicReg, Lrnr_arima, Lrnr_bartMachine, Lrnr_base, Lrnr_bayesglm, Lrnr_bilstm, Lrnr_caret, Lrnr_cv_selector, Lrnr_cv, Lrnr_dbarts, Lrnr_define_interactions, Lrnr_density_discretize, Lrnr_density_hse, Lrnr_density_semiparametric, Lrnr_earth, Lrnr_expSmooth, Lrnr_gam, Lrnr_ga, Lrnr_gbm, Lrnr_glm_fast, Lrnr_glmnet, Lrnr_glm, Lrnr_grf, Lrnr_gru_keras, Lrnr_gts, Lrnr_h2o_grid, Lrnr_hal9001, Lrnr_haldensify, Lrnr_hts, Lrnr_independent_binomial, Lrnr_lightgbm, Lrnr_lstm_keras, Lrnr_mean, Lrnr_multiple_ts, Lrnr_multivariate, Lrnr_nnet, Lrnr_nnls, Lrnr_optim, Lrnr_pca, Lrnr_pkg_SuperLearner, Lrnr_polspline, Lrnr_pooled_hazards, Lrnr_randomForest, Lrnr_ranger, Lrnr_revere_task, Lrnr_rpart, Lrnr_rugarch, Lrnr_screener_augment, Lrnr_screener_coefs, Lrnr_screener_correlation, Lrnr_sl, Lrnr_solnp_density, Lrnr_solnp, Lrnr_stratified, Lrnr_subset_covariates, Lrnr_svm, Lrnr_tsDyn, Lrnr_ts_weights, Lrnr_xgboost, Pipeline, Stack, define_h2o_X(), undocumented_learner

data(mtcars)
mtcars_task <- sl3_Task$new(
  data = mtcars,
  covariates = c(
    "cyl", "disp", "hp", "drat", "wt", "qsec", "vs", "am",
    "gear", "carb"
  ),
  outcome = "mpg"
)
glm_lrnr <- make_learner(Lrnr_glm)

# screening based on \code{\link{Lrnr_ranger}} variable importance
ranger_lrnr_importance <- Lrnr_ranger$new(importance = "impurity_corrected")
ranger_importance_screener <- Lrnr_screener_importance$new(
  learner = ranger_lrnr_importance, num_screen = 3
)
ranger_screen_glm_pipe <- Pipeline$new(ranger_importance_screener, glm_lrnr)
ranger_screen_glm_pipe_fit <- ranger_screen_glm_pipe$train(mtcars_task)

# screening based on \code{\link{Lrnr_randomForest}} variable importance
rf_lrnr <- Lrnr_randomForest$new()
rf_importance_screener <- Lrnr_screener_importance$new(
  learner = rf_lrnr, num_screen = 3
)
rf_screen_glm_pipe <- Pipeline$new(rf_importance_screener, glm_lrnr)
rf_screen_glm_pipe_fit <- rf_screen_glm_pipe$train(mtcars_task)

# screening based on \code{\link{Lrnr_randomForest}} variable importance
xgb_lrnr <- Lrnr_xgboost$new()
xgb_importance_screener <- Lrnr_screener_importance$new(
  learner = xgb_lrnr, num_screen = 3
)
xgb_screen_glm_pipe <- Pipeline$new(xgb_importance_screener, glm_lrnr)
xgb_screen_glm_pipe_fit <- xgb_screen_glm_pipe$train(mtcars_task)