BlockForest Classification Learner

Random forests for blocks of clinical and omics covariate data. Calls blockForest::blockfor() from package blockForest.

In this learner, only the trained forest object ($forest) is retained. The optimized block-specific tuning parameters (paramvalues) and the biased OOB error estimate (biased_oob_error_donotuse) are discarded, as they are either not needed for downstream use or not reliable for performance estimation.

Initial parameter values

• num.threads is initialized to 1 to avoid conflicts with parallelization via future.

Dictionary

This Learner can be instantiated via lrn():

lrn("classif.blockforest")

Meta Information

• Task type: “classif”

• Predict Types: “response”, “prob”

• Feature Types: “logical”, “integer”, “numeric”, “factor”, “ordered”

• Required Packages: mlr3, mlr3extralearners, blockForest

Parameters

Id	Type	Default	Levels	Range
blocks	untyped	-		-
block.method	character	BlockForest	BlockForest, RandomBlock, BlockVarSel, VarProb, SplitWeights	-
num.trees	integer	2000		\([1, \infty)\)
mtry	untyped	NULL		-
nsets	integer	300		\([1, \infty)\)
num.trees.pre	integer	1500		\([1, \infty)\)
splitrule	character	extratrees	extratrees, gini	-
always.select.block	integer	0		\([0, 1]\)
importance	character	-	none, impurity, impurity_corrected, permutation	-
num.threads	integer	-		\([1, \infty)\)
seed	integer	NULL		\((-\infty, \infty)\)
verbose	logical	TRUE	TRUE, FALSE	-

References

Hornung, R., Wright, N. M (2019). “Block Forests: Random forests for blocks of clinical and omics covariate data.” BMC Bioinformatics, 20(1), 1–17. doi:10.1186/s12859-019-2942-y . https://doi.org/10.1186/s12859-019-2942-y.

See also

• Dictionary of Learners: mlr3::mlr_learners.

• as.data.table(mlr_learners) for a table of available Learners in the running session (depending on the loaded packages).

• Chapter in the mlr3book: https://mlr3book.mlr-org.com/chapters/chapter2/data_and_basic_modeling.html#sec-learners

• mlr3learners for a selection of recommended learners.

• mlr3cluster for unsupervised clustering learners.

• mlr3pipelines to combine learners with pre- and postprocessing steps.

• mlr3tuning for tuning of hyperparameters, mlr3tuningspaces for established default tuning spaces.

Author

bblodfon

Super classes

mlr3::Learner -> mlr3::LearnerClassif -> LearnerClassifBlockForest

Methods

Public methods

• LearnerClassifBlockForest$new()

• LearnerClassifBlockForest$importance()

• LearnerClassifBlockForest$clone()

Inherited methods

• mlr3::Learner$base_learner()
• mlr3::Learner$configure()
• mlr3::Learner$encapsulate()
• mlr3::Learner$format()
• mlr3::Learner$help()
• mlr3::Learner$predict()
• mlr3::Learner$predict_newdata()
• mlr3::Learner$print()
• mlr3::Learner$reset()
• mlr3::Learner$selected_features()
• mlr3::Learner$train()
• mlr3::LearnerClassif$predict_newdata_fast()

---

LearnerClassifBlockForest$new()

Creates a new instance of this R6 class.

Usage

LearnerClassifBlockForest$new()

---

LearnerClassifBlockForest$importance()

The importance scores are extracted from the model slot variable.importance.

Usage

LearnerClassifBlockForest$importance()

Returns

Named numeric().

---

LearnerClassifBlockForest$clone()

The objects of this class are cloneable with this method.

Usage

LearnerClassifBlockForest$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

# Define a Task
task = tsk("sonar")

# Create train and test set
ids = partition(task)

# check task's features
task$feature_names
#>  [1] "V1"  "V10" "V11" "V12" "V13" "V14" "V15" "V16" "V17" "V18" "V19" "V2" 
#> [13] "V20" "V21" "V22" "V23" "V24" "V25" "V26" "V27" "V28" "V29" "V3"  "V30"
#> [25] "V31" "V32" "V33" "V34" "V35" "V36" "V37" "V38" "V39" "V4"  "V40" "V41"
#> [37] "V42" "V43" "V44" "V45" "V46" "V47" "V48" "V49" "V5"  "V50" "V51" "V52"
#> [49] "V53" "V54" "V55" "V56" "V57" "V58" "V59" "V6"  "V60" "V7"  "V8"  "V9" 

# partition features to 2 blocks
blocks = list(bl1 = 1:42, bl2 = 43:60)

# define learner
learner = lrn("classif.blockforest", blocks = blocks,
              importance = "permutation", nsets = 10, predict_type = "prob",
              num.trees = 50, num.trees.pre = 10, splitrule = "gini")

# Train the learner on the training ids
learner$train(task, row_ids = ids$train)

# feature importance
learner$importance()
#>           V11            V9           V37           V36           V12 
#>  2.456745e-02  1.660460e-02  1.617900e-02  1.515646e-02  1.489175e-02 
#>           V45           V48           V19           V16           V15 
#>  9.742701e-03  6.658790e-03  5.920587e-03  5.430257e-03  5.063309e-03 
#>            V8           V51            V2           V55           V21 
#>  4.498025e-03  4.494935e-03  4.302920e-03  4.271657e-03  4.100458e-03 
#>           V22           V20           V23           V13           V44 
#>  3.318493e-03  3.302800e-03  3.252842e-03  3.238761e-03  3.235585e-03 
#>           V33           V10           V26            V6            V3 
#>  3.180045e-03  2.991630e-03  2.689020e-03  2.638000e-03  2.607440e-03 
#>           V14           V27            V1           V31            V7 
#>  2.568023e-03  2.518262e-03  2.396104e-03  2.384569e-03  2.327401e-03 
#>           V34           V52           V47           V49           V29 
#>  2.184355e-03  2.013919e-03  1.840835e-03  1.748582e-03  1.656139e-03 
#>           V32           V18           V58           V24           V54 
#>  1.400262e-03  1.084261e-03  1.073745e-03  9.479697e-04  4.546649e-04 
#>           V50           V42           V28           V30           V53 
#>  2.452651e-04  2.153658e-04  2.209888e-05  4.094810e-06 -2.508284e-05 
#>           V57            V4           V46           V40           V25 
#> -1.095593e-04 -1.419246e-04 -1.451425e-04 -7.015111e-04 -7.021957e-04 
#>           V17           V59            V5           V43           V56 
#> -7.687888e-04 -8.917972e-04 -9.091277e-04 -9.110489e-04 -9.974393e-04 
#>           V39           V41           V38           V35           V60 
#> -1.001701e-03 -1.375581e-03 -1.440435e-03 -1.539927e-03 -4.054114e-03 

# Make predictions for the test observations
pred = learner$predict(task, row_ids = ids$test)
pred
#> 
#> ── <PredictionClassif> for 69 observations: ────────────────────────────────────
#>  row_ids truth response    prob.M    prob.R
#>        3     R        M 0.6472540 0.3527460
#>        7     R        M 0.5119524 0.4880476
#>       10     R        R 0.3213492 0.6786508
#>      ---   ---      ---       ---       ---
#>      200     M        M 0.7712778 0.2287222
#>      205     M        M 0.5553730 0.4446270
#>      207     M        M 0.6830952 0.3169048

# Score the predictions
pred$score()
#> classif.ce 
#>  0.2318841