tune_calibratedva: Chooses parameter values for calibration using WAIC
In jfiksel/CalibratedVA: CalibratedVA

tune_calibratedva

R Documentation

Chooses parameter values for calibration using WAIC

Description

Uses a grid search to choose the shrinkage parameter for calibration which gives the lowest WAIC values.

Usage

tune_calibratedva(
  va_unlabeled,
  va_labeled = NULL,
  gold_standard = NULL,
  causes,
  method = c("mshrink", "pshrink"),
  alpha_vec = NULL,
  lambda_vec = NULL,
  samples_list = NULL,
  which.multimodal = "all",
  which.rhat = "all",
  ...
)

Arguments

`va_unlabeled`	A matrix or list of causes for individuals without labeled causes of death. See `calibratedva`.
`va_labeled`	A matrix or list of causes for individuals with labeled causes of death. See `calibratedva`.
`gold_standard`	A matrix where each row represents either the true cause for an individual with a labeled cause of death. See `calibratedva`.
`causes`	A character vector with the names of the causes. These should correspond to the columns of `A_U`, `A_L`, and `G_L`
`method`	One of either "mshrink" (default) for M-shrinkage or "pshrink" for p-shrinkage
`alpha_vec`	If using M-shrinkage vector of values for alpha for which the function will evaluate the WAIC. If not provided, the function will auto-generate a vector.
`lambda_vec`	If using p-shrinkage vector of values for lambda for which the function will evaluate the WAIC. If not provided, the function will auto-generate a vector.
`which.multimodal`	A character specifying whether both p and M (which.multimodal = "all") should be evaluated for multimodality, or just p (which.multimodal = "p")
`which.rhat`	A character specifying whether both p and M (which.rhat = "all") should be evaluated for convergence, or just p (which.rhat = "p")
`...`	Additional arguments passed into `calibratedva`.

Value

A list with the following components.

final_model: A list with output specified in calibratedva object containing the posterior samples for the best value of the parameter (either alpha or lambda)
alpha_final: The chosen value of alpha for the posterior samples. Null if using p-shrinkage.
lambda_final: The chosen value of lambda for the posterior samples. Null if using M-shrinkage.
waic_df: A data frame containing a row for each parameter evaluated, which gives the WAIC, whether or not the posteriors were multimodal, and the Rhat of the posterior samples.