inferEvidence2: inferEvidence2

In oyvble/MPSproto: A Quantitative Model for MPS data with complex stutters

inferEvidence2

Description

Function to perform inference of new evidence (Penalized MLE based)

Usage

inferEvidence2(
  mlefit,
  calibration = NULL,
  priorAsLN = FALSE,
  reOpt = FALSE,
  nDone = 3,
  delta = 1,
  steptol = 1e-04,
  seed = NULL,
  verbose = FALSE
)

Arguments

mlefit	Output object from inferEvidence function
calibration	A list from calibration results (per marker). Must contain the following elements (in following order): AT Analytical threshold markerEff Marker efficiency param (possibly also SD if running inferEvidence2)
priorAsLN	Whether model prior of marker efficiency as a log-normal (SD from logged values)
reOpt	Whether to re-run optimization with random start values
nDone	Number of optimizations required providing equivalent results (same logLik value obtained)
delta	Scaling of variation of normal distribution when drawing random startpoints. Default is 1.
steptol	Argument used in the nlm function for faster return from the optimization (tradeoff is lower accuracy).
seed	The user can set seed if wanted
verbose	Whether to print out progress

Details

Optimizes sample specific params (mx,mu,omega, Am). Using MLE as initial params for optimization. Marker efficiency parameter is penalized on prior.

oyvble/MPSproto documentation built on March 19, 2024, 5:32 a.m.