PatientLevelPrediction (version 6.4.1)

setRandomForest: Create setting for random forest model using sklearn

Description

Create setting for random forest model using sklearn

Usage

setRandomForest(
  ntrees = list(100, 500),
  criterion = list("gini"),
  maxDepth = list(4, 10, 17),
  minSamplesSplit = list(2, 5),
  minSamplesLeaf = list(1, 10),
  minWeightFractionLeaf = list(0),
  mtries = list("sqrt", "log2"),
  maxLeafNodes = list(NULL),
  minImpurityDecrease = list(0),
  bootstrap = list(TRUE),
  maxSamples = list(NULL, 0.9),
  oobScore = list(FALSE),
  nJobs = list(NULL),
  classWeight = list(NULL),
  seed = sample(1e+05, 1)
)

Value

a modelSettings object

Arguments

ntrees

(list) The number of trees to build

criterion

(list) The function to measure the quality of a split. Supported criteria are “gini” for the Gini impurity and “entropy” for the information gain. Note: this parameter is tree-specific.

maxDepth

(list) The maximum depth of the tree. If NULL, then nodes are expanded until all leaves are pure or until all leaves contain less than minSamplesSplit samples.

minSamplesSplit

(list) The minimum number of samples required to split an internal node

minSamplesLeaf

(list) The minimum number of samples required to be at a leaf node. A split point at any depth will only be considered if it leaves at least minSamplesLeaf training samples in each of the left and right branches. This may have the effect of smoothing the model, especially in regression.

minWeightFractionLeaf

(list) The minimum weighted fraction of the sum total of weights (of all the input samples) required to be at a leaf node. Samples have equal weight when sampleWeight is not provided.

mtries

(list) The number of features to consider when looking for the best split:

• int then consider max_features features at each split.

• float then max_features is a fraction and round(max_features * n_features) features are considered at each split

• 'sqrt' then max_features=sqrt(n_features)

• 'log2' then max_features=log2(n_features)

• NULL then max_features=n_features

maxLeafNodes

(list) Grow trees with max_leaf_nodes in best-first fashion. Best nodes are defined as relative reduction in impurity. If None then unlimited number of leaf nodes.

minImpurityDecrease

(list) A node will be split if this split induces a decrease of the impurity greater than or equal to this value.

bootstrap

(list) Whether bootstrap samples are used when building trees. If False, the whole dataset is used to build each tree.

maxSamples

(list) If bootstrap is True, the number of samples to draw from X to train each base estimator.

oobScore

(list) Whether to use out-of-bag samples to estimate the generalization score. Only available if bootstrap=True.

nJobs

The number of jobs to run in parallel.

classWeight

(list) Weights associated with classes. If not given, all classes are supposed to have weight one. NULL, “balanced”, “balanced_subsample”

seed

A seed when training the final model

Examples

 # dontrun reason: requires python's scikit-learn, checkSklearn() will error without it 
if (FALSE) { 
plpData <- getEunomiaPlpData()
model <- setRandomForest(ntrees = list(100),
                          maxDepth = list(4),
                          minSamplesSplit = list(2),
                          minSamplesLeaf = list(10),
                          maxSamples = list(0.9),
                          seed = 42)
saveLoc <- file.path(tempdir(), "randomForest")
results <- runPlp(plpData, modelSettings = model, saveDirectory = saveLoc)
# clean up
unlink(saveLoc, recursive = TRUE)
}