check_early_stopping_condition: Check if the early stopping condition is satisfied

In junyangq/multiSnpnet: Fit a sparse reduced rank regression model on large-scale SNP data and multivariate responses

Check if the early stopping condition is satisfied

Description

For the specified traits, we check if the validation set metric (and validation set AUC) has been decreasing in the last two steps. If also check if the average of the validation set metric and the validation set AUC has been decreasing in the last two steps. If no traits were provided, we check it for all the traits considered in SRRR.

Usage

check_early_stopping_condition(
  ilam,
  metric_val,
  AUC_val = NULL,
  traits = NULL,
  weight = NULL,
  check_average = TRUE,
  stopping.lag = 2
)

Arguments

ilam	Lambda index
metric_val	Validation set metric
AUC_val	Validation set AUC
traits	(optional) subset of traits
weight	weights (named list) for the weighted average computation.
check_average	whether to check the average of the metrics
stopping.lag	how many iterations shall we wait after the argmax until we stop

junyangq/multiSnpnet documentation built on Oct. 19, 2023, 8:22 p.m.