restartTuneRanger: restartTuneRanger

In PhilippPro/tuneRF: Tune Random Forest of the 'ranger' Package

restartTuneRanger

Description

Restarts the tuning process if an error occured.

Usage

restartTuneRanger(save.file.path = "optpath.RData", task, measure = NULL)

Arguments

save.file.path	File name in the current working directory to which interim results were saved by tuneRanger.
task	The mlr task created by makeClassifTask or makeRegrTask.
measure	Performance measure that was already used in the original tuneRanger process.

Value

A list with elements

recommended.pars	Recommended hyperparameters.
results	A data.frame with all evaluated hyperparameters and performance and time results for each run.

No model is build.

Examples

## Not run: 
library(tuneRanger)
library(mlr)

# iris is a bit nonsense here
# A mlr task has to be created in order to use the package
# the already existing iris task is used here
estimateTimeTuneRanger(iris.task)
# temporarily file name to save results
path = tempfile()
res = tuneRanger(iris.task, measure = list(multiclass.brier), num.trees = 1000, 
  num.threads = 8, iters = 70, save.file.path = path)

# Mean of best 5 % of the results
res

# Restart after failing in one of the iterations:
res = restartTuneRanger(save.file.path = path, iris.task, 
measure = list(multiclass.brier))
## End(Not run)

PhilippPro/tuneRF documentation built on March 24, 2024, 1:23 p.m.