HMDA (version 0.1.1)

hmda.adjust.params: Adjust Hyperparameter Combinations

Description

This internal function prunes or expands a list of hyperparameters so that the total number of model combinations, computed as the product of the lengths of each parameter vector, is near the desired target (n_models). It first prunes the parameter with the largest number of values until the product is less than or equal to n_models. Then, if the product is much lower than the target (less than half of n_models), it attempts to expand the parameter with the smallest number of values by adding a midpoint value (if numeric).

Usage

hmda.adjust.params(params, n_models)

Value

A list of hyperparameter vectors that has been pruned or expanded so that the product of their lengths is near

n_models.

Arguments

params

A list of hyperparameter vectors.

n_models

Integer. The desired target number of model combinations.

Author

E. F. Haghish

Details

The function calculates the current product of the lengths of the hyperparameter vectors. In a loop, it removes the last element from the parameter vector with the largest length until the product is less than or equal to n_models. If the resulting product is less than half of n_models, the function attempts to expand the parameter with the smallest length by computing a midpoint between the two closest numeric values. The expansion stops if no new value can be added, to avoid an infinite loop.

Examples

  # Example 1: Adjust a hyperparameter grid for 100 models.
  params <- list(
    alpha = c(0.1, 0.2, 0.3, 0.4),
    beta = c(1, 2, 3, 4, 5),
    gamma = c(10, 20, 30)
  )
  new_params <- hmda.adjust.params(params, n_models = 100)
  print(new_params)

  # Example 2: The generated hyperparameters range between min and max of each
  # vector in the list
  params <- list(
    alpha = c(0.1, 0.2),
    beta = c(1, 2, 3),
    gamma = c(10, 20)
  )
  new_params <- hmda.adjust.params(params, n_models = 1000)
  print(new_params)