assign.convergence: Check the convergence of the MCMC chain

In compbiomed/ASSIGN: Adaptive Signature Selection and InteGratioN (ASSIGN)

Check the convergence of the MCMC chain

Description

The assign.convergence checks the convergence of the MCMC chain of the model parameters generated by the Gibbs sampling algorithm.

Usage

assign.convergence(
  test,
  burn_in = 0,
  iter = 2000,
  parameter = c("B", "S", "Delta", "beta", "kappa", "gamma", "sigma"),
  whichGene,
  whichSample,
  whichPath
)

Arguments

test	The list object returned from the assign.mcmc function. The list components are the MCMC chains of the B, S, Delta, beta, gamma, and sigma.
burn_in	The number of burn-in iterations. These iterations are discarded when computing the posterior means of the model parameters. The default is 0.
iter	The number of total iterations. The default is 2000.
parameter	A character string indicating which model parameter is to be checked for convergence. This must be one of "B", "S", "Delta", "beta", "kappa", "gamma", and "sigma".
whichGene	A numerical value indicating which gene is to be checked for convergence. The value has to be in the range between 1 and G.
whichSample	A numerical value indicating which test sample is to be checked for convergence. The value has to be in the range between 1 and N.
whichPath	A numerical value indicating which pathway is to be checked for convergence. The value has to be in the range between 1 and K.

Details

To compute the convergence of the gth gene in B, set whichGene=g, whichSample=NA, whichPath=NA.

To compute the convergence of the gth gene in the kth pathway within the signature matrix (S), set whichGene=g, whichSample=NA, whichPath=NA.

To compute the convergence of the kth pathway in the jth test sample within the pathway activation matrix (A), set whichGene=NA, whichSample=n, whichPath=k.

Value

The assign.convergence function returns the a vector of the estimated values from each Gibbs sampling iteration of the model parameter to be checked, and a trace plot of this parameter.

Author(s)

Ying Shen

Examples

## Not run: 
# check the 10th gene in the 1st pathway for the convergence
trace.plot <- assign.convergence(test=mcmc.chain, burn_in=0, iter=2000,
                                 parameter="S", whichGene=10, whichSample=NA,
                                 whichPath=1)

## End(Not run)

compbiomed/ASSIGN documentation built on June 28, 2023, 4 a.m.