Preprocess a dataframe of covariate values, converting categorical variables to integers and one-hot encoding if need be. Returns a list including a matrix of preprocessed covariate values and associated tracking.

Preprocess a dataframe of covariate values, converting categorical variables to integers and one-hot encoding if need be. Returns a list including a matrix of preprocessed covariate values and associated tracking.

Usage

createForestCovariates(
  input_data,
  ordered_cat_vars = NULL,
  unordered_cat_vars = NULL
)

Arguments

input_data

Dataframe or matrix of covariates. Users may pre-process any categorical variables as factors but it is not necessary.

ordered_cat_vars

(Optional) Vector of names of ordered categorical variables, or vector of column indices if input_data is a matrix.

unordered_cat_vars

(Optional) Vector of names of unordered categorical variables, or vector of column indices if input_data is a matrix.

Value

List with preprocessed data and details on the number of each type of variable, unique categories associated with categorical variables, and the vector of feature types needed for calls to BART and BCF.

Examples

cov_df <- data.frame(x1 = 1:5, x2 = 5:1, x3 = 6:10)
preprocess_list <- createForestCovariates(cov_df)
X <- preprocess_list$X