ranger: Ranger

Description

Ranger is a fast implementation of Random Forest (Breiman 2001) or recursive partitioning, particularly suited for high dimensional data. Classification, regression, and survival forests are supported. Classification and regression forests are implemented as in the original Random Forest (Breiman 2001), survival forests as in Random Survival Forests (Ishwaran et al. 2008).

Usage

ranger(formula = NULL, data = NULL, num.trees = 500, mtry = NULL, importance = "none", write.forest = FALSE, probability = FALSE, min.node.size = NULL, replace = TRUE, sample.fraction = ifelse(replace, 1, 0.632), case.weights = NULL, splitrule = NULL, alpha = 0.5, minprop = 0.1, split.select.weights = NULL, always.split.variables = NULL, respect.unordered.factors = "ignore", scale.permutation.importance = FALSE, keep.inbag = FALSE, holdout = FALSE, num.threads = NULL, save.memory = FALSE, verbose = TRUE, seed = NULL, dependent.variable.name = NULL, status.variable.name = NULL, classification = NULL)

Arguments

formula

Object of class formula or character describing the model to fit.

data

Training data of class data.frame, matrix or gwaa.data (GenABEL).

num.trees

Number of trees.

mtry

Number of variables to possibly split at in each node. Default is the (rounded down) square root of the number variables.

importance

Variable importance mode, one of 'none', 'impurity', 'permutation'. The 'impurity' measure is the Gini index for classification and the variance of the responses for regression. For survival, only 'permutation' is available.

write.forest

Save ranger.forest object, needed for prediction.

probability

Grow a probability forest as in Malley et al. (2012).

min.node.size

Minimal node size. Default 1 for classification, 5 for regression, 3 for survival, and 10 for probability.

replace

Sample with replacement.

sample.fraction

Fraction of observations to sample. Default is 1 for sampling with replacement and 0.632 for sampling without replacement.

case.weights

Weights for sampling of training observations. Observations with larger weights will be selected with higher probability in the bootstrap (or subsampled) samples for the trees.

splitrule

Splitting rule, regressiond and survival only. For regression one of "variance" or "maxstat" with default "variance". For survival "logrank", "C" or "maxstat" with default "logrank".

alpha

For "maxstat" splitrule: Significance threshold to allow splitting.

minprop

For "maxstat" splitrule: Lower quantile of covariate distribtuion to be considered for splitting.

split.select.weights

Numeric vector with weights between 0 and 1, representing the probability to select variables for splitting. Alternatively, a list of size num.trees, containing split select weight vectors for each tree can be used.

always.split.variables

Character vector with variable names to be always tried for splitting.

respect.unordered.factors

Handling of unordered factor covariates. One of 'ignore', 'order' and 'partition' with default 'ignore'. Alternatively TRUE (='order') or FALSE (='ignore') can be used. See below for details.

scale.permutation.importance

Scale permutation importance by standard error as in (Breiman 2001). Only applicable if permutation variable importance mode selected.

keep.inbag

Save how often observations are in-bag in each tree.

holdout

Hold-out mode. Hold-out all samples with case weight 0 and use these for variable importance and prediction error.

num.threads

Number of threads. Default is number of CPUs available.

save.memory

Use memory saving (but slower) splitting mode. No effect for GWAS data.

verbose

Verbose output on or off.

seed

Random seed. Default is NULL, which generates the seed from R.

dependent.variable.name

Name of dependent variable, needed if no formula given. For survival forests this is the time variable.

status.variable.name

Name of status variable, only applicable to survival data and needed if no formula given. Use 1 for event and 0 for censoring.

classification

Only needed if data is a matrix. Set to TRUE to grow a classification forest.

Value

forest: Saved forest (If write.forest set to TRUE). Note that the variable IDs in the split.varIDs object do not necessarily represent the column number in R.
predictions: Predicted classes/values, based on out of bag samples (classification and regression only).
forest: Saved forest (If write.forest set to TRUE). Note that the variable IDs in the split.varIDs object do not necessarily represent the column number in R.
predictions: Predicted classes/values, based on out of bag samples (classification and regression only).
variable.importance: Variable importance for each independent variable.
prediction.error: Overall out of bag prediction error. For classification this is the fraction of missclassified samples, for regression the mean squared error and for survival one minus Harrell's c-index.
r.squared: R squared. Also called explained variance or coefficient of determination (regression only).
confusion.matrix: Contingency table for classes and predictions based on out of bag samples (classification only).
unique.death.times: Unique death times (survival only).
chf: Estimated cumulative hazard function for each sample (survival only).
survival: Estimated survival function for each sample (survival only).
call: Function call.
num.trees: Number of trees.
num.independent.variables: Number of independent variables.
mtry: Value of mtry used.
min.node.size: Value of minimal node size used.
treetype: Type of forest/tree. classification, regression or survival.
importance.mode: Importance mode used.
num.samples: Number of samples.
inbag.counts: Number of times the observations are in-bag in the trees.

Details

The tree type is determined by the type of the dependent variable. For factors classification trees are grown, for numeric values regression trees and for survival objects survival trees. The Gini index is used as splitting rule for classification. For regression, the estimated response variances or maximally selected rank statistics (Wright et al. 2016) can be used. For Survival the log-rank test, a C-index based splitting rule (Schmid et al. 2015) and maximally selected rank statistics (Wright et al. 2016) are available.

With the probability option and factor dependent variable a probability forest is grown. Here, the node impurity is used for splitting, as in classification forests. Predictions are class probabilities for each sample. In contrast to other implementations, each tree returns a probability estimate and these estimates are averaged for the forest probability estimate. For details see Malley et al. (2012).

Note that for classification and regression nodes with size smaller than min.node.size can occur, like in original Random Forest. For survival all nodes contain at least min.node.size samples. Variables selected with always.split.variables are tried additionaly to the mtry variables randomly selected. In split.select.weights variables weighted with 0 are never selected and variables with 1 are always selected. Weights do not need to sum up to 1, they will be normalized later. The usage of split.select.weights can increase the computation times for large forests.

Unordered factor covariates can be handled in 3 different ways by using respect.unordered.factors: For 'ignore' all factors are regarded ordered, for 'partition' all possible 2-partitions are considered for splitting. For 'order' and 2-class classification the factor levels are ordered by their proportion falling in the second class, for regression by their mean response, as described in Hastie et al. (2009), chapter 9.2.4. For multiclass classification and survival outcomes, 'order' is experimental and should be used with care. The use of 'order' is recommended for 2-class classification and regression, as it computationally fast and can handle an unlimited number of factor levels. Note that the factors are only reordered once and not again in each split.

For a large number of variables and data frame as input data the formula interface can be slow or impossible to use. Alternatively dependent.variable.name (and status.variable.name for survival) can be used. Consider setting save.memory = TRUE if you encounter memory problems for very large datasets.

For GWAS data consider combining ranger with the GenABEL package. See the Examples section below for a demonstration using Plink data. All SNPs in the GenABEL object will be used for splitting. To use only the SNPs without sex or other covariates from the phenotype file, use 0 on the right hand side of the formula. Note that missing values are treated as an extra category while splitting.

See https://github.com/imbs-hl/ranger for the development version.

To use multithreading on Microsoft Windows platforms, there are currently two options:

Use R-devel with the new toolchain
Install a binary version of ranger, download: https://github.com/imbs-hl/ranger/releases

References

Wright, M. N. & Ziegler, A. (2016). ranger: A Fast Implementation of Random Forests for High Dimensional Data in C++ and R. Journal of Statistical Software, in press. http://arxiv.org/abs/1508.04409.
Schmid, M., Wright, M. N. & Ziegler, A. (2015). On the Use of Harrell's C for Node Splitting in Random Survival Forests. Technical Report. http://arxiv.org/abs/1507.03092.
Wright, M. N., Dankowski, T. & Ziegler, A. (2016). Random forests for survival analysis using maximally selected rank statistics. Technical Report. http://arxiv.org/abs/1605.03391.
Breiman, L. (2001). Random forests. Mach Learn, 45(1), 5-32.
Ishwaran, H., Kogalur, U. B., Blackstone, E. H., & Lauer, M. S. (2008). Random survival forests. Ann Appl Stat, 841-860.
Malley, J. D., Kruppa, J., Dasgupta, A., Malley, K. G., & Ziegler, A. (2012). Probability machines: consistent probability estimation using nonparametric learning machines. Methods Inf Med, 51(1), 74.
Hastie, T., Tibshirani, R., Friedman, J. (2009). The Elements of Statistical Learning. Springer, New York. 2nd edition.

Examples

Run this code

require(ranger)

## Classification forest with default settings
ranger(Species ~ ., data = iris)

## Prediction
train.idx <- sample(nrow(iris), 2/3 * nrow(iris))
iris.train <- iris[train.idx, ]
iris.test <- iris[-train.idx, ]
rg.iris <- ranger(Species ~ ., data = iris.train, write.forest = TRUE)
pred.iris <- predict(rg.iris, dat = iris.test)
table(iris.test$Species, pred.iris$predictions)

## Variable importance
rg.iris <- ranger(Species ~ ., data = iris, importance = "impurity")
rg.iris$variable.importance

## Survival forest
require(survival)
rg.veteran <- ranger(Surv(time, status) ~ ., data = veteran)
plot(rg.veteran$unique.death.times, rg.veteran$survival[1,])

## Alternative interface
ranger(dependent.variable.name = "Species", data = iris)

## Not run: 
# ## Use GenABEL interface to read Plink data into R and grow a classification forest
# ## The ped and map files are not included
# library(GenABEL)
# convert.snp.ped("data.ped", "data.map", "data.raw")
# dat.gwaa <- load.gwaa.data("data.pheno", "data.raw")
# phdata(dat.gwaa)$trait <- factor(phdata(dat.gwaa)$trait)
# ranger(trait ~ ., data = dat.gwaa)
# ## End(Not run)

Run the code above in your browser using DataLab