Generate Resampled Model Inputs — generate_model

Reads the engineered feature layer, prepares a base tidymodels recipe, and generates resampled datasets (Baseline, Under, SMOTE, Adasyn, Tomek) across all months, saving them to the 05_model_input prefix.

Usage

generate_model_inputs(
  feature_prefix = "baf-fraud/04_feature/variant=Base",
  out_prefix = "baf-fraud/05_model_input",
  bucket_name = "lake"
)

Arguments

feature_prefix: Character. Input prefix (e.g., "04_feature/variant=Base").
out_prefix: Character. Output prefix base (e.g., "05_model_input").
bucket_name: Character. Bucket name. Default "lake".

Value

Character. The output prefix (for targets dependency tracking).