Merges a list of forests that were grown using the same data into one large forest.
Merges a list of forests that were grown using the same data into one large forest.
merge_forests(forest_list, compute.oob.predictions = TRUE)
forest_list |
A 'list' of forests to be concatenated. All forests must be of the same type, and the type must be a subclass of 'grf'. In addition, all forests must have the same 'ci.group.size'. Other tuning parameters (e.g. alpha, mtry, min.node.size, imbalance.penalty) are allowed to differ across forests. |
compute.oob.predictions |
Whether OOB predictions on training set should be precomputed. Note that even if OOB predictions have already been precomputed for the forests in 'forest_list', those predictions are not used. Instead, a new set of oob predictions is computed anew using the larger forest. Default is TRUE. |
A single forest containing all the trees in each forest in the input list.
# Train standard regression forests n <- 50 p <- 10 X <- matrix(rnorm(n * p), n, p) Y <- X[, 1] * rnorm(n) r.forest1 <- regression_forest(X, Y, compute.oob.predictions = FALSE, num.trees = 100) r.forest2 <- regression_forest(X, Y, compute.oob.predictions = FALSE, num.trees = 100) # Join the forests together. The resulting forest will contain 200 trees. big_rf <- merge_forests(list(r.forest1, r.forest2))
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.