Preprocess data for SPQR/SPQRX modeling — preprocessing.data • SPQRX

Splits data into training, validation, and testing sets and optionally applies normalization to predictors and response.

Usage

preprocessing.data(
  x,
  y,
  n.knots,
  testing_ratio = 0.1,
  valid_ratio = 0.1,
  normalize = FALSE
)

Arguments

x: Numeric matrix of covariates.
y: Numeric vector or matrix of response values.
n.knots: Integer. Number of spline knots.
testing_ratio: Proportion of data allocated to testing.
valid_ratio: Proportion of training data allocated to validation.
normalize: Logical. If TRUE, standardizes predictors and rescales response.

Value

A list containing:

Training, validation, and testing splits
Spline knot locations
Normalization metadata (if applicable)

Details

When normalization is enabled, predictors are standardized using training-set statistics, and the response is rescaled to the unit interval.