Split Data into Training and Testing Sets

Divides dataset into training and testing sets using random sampling. Maintains data integrity for both binomial and survival analysis types.

Usage

SplitTrainTest(x, y, train_ratio, type = c("binomial", "survival"), seed)

Arguments

x

Predictor matrix or data frame.

y

Outcome vector (binomial) or matrix with time/status (survival).

train_ratio

Proportion for training (0-1). Default is `0.7`.

type

Analysis type: `"binomial"` or `"survival"`.

seed

Random seed for reproducibility.

Value

List containing:

train.x

Training predictors matrix

train.y

Training outcomes

test.x

Testing predictors matrix

test.y

Testing outcomes

train_sample

Indices of training samples

Examples

data_matrix <- matrix(rnorm(200), ncol = 2)
outcome_vector <- rbinom(100, 1, 0.5)
split_data <- SplitTrainTest(
  data_matrix, outcome_vector,
  train_ratio = 0.7,
  type = "binomial", seed = 123
)