Runs ranger for a vector of mtry values and returns the out of bag prediction error for each run.
For each mtry the analysis will be replicated n times.
Usage
tune.mtry(mtry, n = 1, ...)
Arguments
mtry
Vector of mtry values to try.
n
Number of replicates.
...
Further arguments passed to Ranger.
Value
Prediction error for each tried mtry value.
Details
The runs of ranger for different values of mtry are NOT based on the same bootstrap sample.
Be sure to run enough replicates (n) before you trust the results.