grf (version 0.10.3)

tune_causal_forest: Causal forest tuning

Description

Finds the optimal parameters to be used in training a regression forest. This method currently tunes over min.node.size, mtry, sample.fraction, alpha, and imbalance.penalty. Please see the method 'causal_forest' for a description of the standard causal forest parameters. Note that if fixed values can be supplied for any of the parameters mentioned above, and in that case, that parameter will not be tuned. For example, if this method is called with min.node.size = 10 and alpha = 0.7, then those parameter values will be treated as fixed, and only sample.fraction and imbalance.penalty will be tuned.

Usage

tune_causal_forest(X, Y, W, Y.hat, W.hat, sample.weights = NULL,
  num.fit.trees = 200, num.fit.reps = 50, num.optimize.reps = 1000,
  min.node.size = NULL, sample.fraction = 0.5, mtry = NULL,
  alpha = NULL, imbalance.penalty = NULL, stabilize.splits = TRUE,
  honesty = TRUE, honesty.fraction = NULL, clusters = NULL,
  samples.per.cluster = NULL, num.threads = NULL, seed = NULL)

Arguments

X

The covariates used in the causal regression.

Y

The outcome.

W

The treatment assignment (may be binary or real).

Y.hat

Estimates of the expected responses E[Y | Xi], marginalizing over treatment. See section 6.1.1 of the GRF paper for further discussion of this quantity.

W.hat

Estimates of the treatment propensities E[W | Xi].

sample.weights

Weights defining the population on which we want our estimator of tau(x) to perform well on average. If NULL, this is the population from which X1 ... Xn are sampled. Otherwise, it is a reweighted version, in which we observe Xi with probability proportional to sample.weights[i].

num.fit.trees

The number of trees in each 'mini forest' used to fit the tuning model.

num.fit.reps

The number of forests used to fit the tuning model.

num.optimize.reps

The number of random parameter values considered when using the model to select the optimal parameters.

min.node.size

A target for the minimum number of observations in each tree leaf. Note that nodes with size smaller than min.node.size can occur, as in the original randomForest package.

sample.fraction

Fraction of the data used to build each tree. Note: If honesty = TRUE, these subsamples will further be cut by a factor of honesty.fraction.

mtry

Number of variables tried for each split.

alpha

A tuning parameter that controls the maximum imbalance of a split.

imbalance.penalty

A tuning parameter that controls how harshly imbalanced splits are penalized.

stabilize.splits

Whether or not the treatment should be taken into account when determining the imbalance of a split (experimental).

honesty

Whether to use honest splitting (i.e., sub-sample splitting).

honesty.fraction

The fraction of data that will be used for determining splits if honesty = TRUE. Corresponds to set J1 in the notation of the paper. When using the defaults (honesty = TRUE and honesty.fraction = NULL), half of the data will be used for determining splits

clusters

Vector of integers or factors specifying which cluster each observation corresponds to.

samples.per.cluster

If sampling by cluster, the number of observations to be sampled from each cluster. Must be less than the size of the smallest cluster. If set to NULL software will set this value to the size of the smallest cluster.#'

num.threads

Number of threads used in training. By default, the number of threads is set to the maximum hardware concurrency.

seed

The seed of the C++ random number generator.

Value

A list consisting of the optimal parameter values ('params') along with their debiased error ('error').

Examples

# NOT RUN {
# Find the optimal tuning parameters.
n = 50; p = 10
X = matrix(rnorm(n*p), n, p)
W = rbinom(n, 1, 0.5)
Y = pmax(X[,1], 0) * W + X[,2] + pmin(X[,3], 0) + rnorm(n)
Y.hat = predict(regression_forest(X, Y))$predictions
W.hat = rep(0.5, n)
params = tune_causal_forest(X, Y, W, Y.hat, W.hat)$params

# Use these parameters to train a regression forest.
tuned.forest = causal_forest(X, Y, W,
    Y.hat = Y.hat, W.hat = W.hat, num.trees = 1000,
    min.node.size = as.numeric(params["min.node.size"]),
    sample.fraction = as.numeric(params["sample.fraction"]),
    mtry = as.numeric(params["mtry"]),
    alpha = as.numeric(params["alpha"]),
    imbalance.penalty = as.numeric(params["imbalance.penalty"])
# }
# NOT RUN {
# }