fit_comixture: Fit shared component gaussian mixed model.

In daniel-wells/comixr: Shared component gaussian mixture models in R

Description

fit_comixture returns parameters for gassians fitted to the data.

Usage

fit_comixture(data, input.parameters, rho = 0.5, max.iterations = 40,
  break.parameter = 5, algorithm = "EM", quiet = FALSE)

Arguments

data	data.frame with two columns, first being numeric read.counts, and second being charachter segment names. Column names can be whatever.
input.parameters	Data frame holding initialisation parameters for the prior distribution parameters. With numeric columns "scale" and "shape" the parameters for the gamma prior on precision; "mean" and "nu" for the mean and precision parameters for the normal prior on the mean; and a charachter column "component.type" with either "common" or "specific"
rho	Numeric value for initial value of rho parameter
max.iterations	maximum number of iterations
break.parameter	Numeric value, when the difference in log likelihood between two iterations is less than this value the iteration will stop
algorithm	Charachter specifying which inference algorithm to use. One of either "EM" (Expectation-Maximisation) or "VB" (Variational-Bayes)
quiet,	Logical specifying whether the iteration number should be printed for each iteration

Value

A list of parameters

Examples

# see vignette

daniel-wells/comixr documentation built on May 14, 2019, 3:39 p.m.