xgboost_fit_predict_r: Adapt xgboost to a fit_predict() signature

In WinVector/CVRTSEncoder: Categorical Variable Residual Trajectory Specteral Encoder

Description

Adapt xgboost to a fit_predict() signature

Usage

xgboost_fit_predict_r(train_data, vars, dep_var, application_data, ...,
  cl = NULL)

Arguments

train_data	data.frame training data.
vars	character vector, explanatory variable names.
dep_var	character, the name of dependent variable.
application_data	data.frame application data
...	not used, force arguments to be bound by name
cl	parallel cluster for processing

Value

a vector with the same number of rows as data representing the model predictions

Examples

data <- iris
vars <- c("Sepal.Length", "Sepal.Width",
          "Petal.Length", "Petal.Width")
data$tvalue <- as.numeric(data$Species=="versicolor")
dep_var <- "tvalue"
xval <- vtreat::kWayStratifiedY(
  nrow(data),
  3,
  data,
  data[[dep_var]])
train_data <- data[xval[[1]]$train, , drop = FALSE]
application_data <- data[xval[[1]]$app, , drop = FALSE]
preds <- xgboost_fit_predict_r(
  train_data = train_data,
  vars = vars,
  dep_var = dep_var,
  application_data = application_data)

WinVector/CVRTSEncoder documentation built on June 7, 2019, 9:53 a.m.