model_select_t_mixture: Select model of a t-mixture distribution

In sp2019-antibiotics/Team-Student: T-Mixture

Description

Computes AIC or BIC for each model (K = 1,...,Kmax). Selects the best model and provides an analysis.

Arguments

n	Frequencies (integer vector of length J; non-negative entries)
j	Possible values (numeric vector of length J > 0)
Kmax	Number of maximum components (numeric scalar)
atoms	Values marking 'resistant' observations (numeric vector of length < J; elements must also be in j)
draw	Should results be visualized? (boolean scalar)
Ecoff.quantile	Which quantile should be used for Ecoff? (numeric scalar within (0, 1))
pi_cutoff	Lower bound for group size of 'wild type' (numeric sclara within (0, 1))
alpha	Hyperparameters for MAP estimation (numeric scalar)
beta	Hyperparameters for MAP estimation (numeric scalar)
memb.exp	Clustering parameter (numeric scalar > 1)
maxiter	Maximum number of iterationsn (integer scalar)
eps	Convergence criterion
optim.method	a

Value

An object of class 'list' with elements:

parameters	Kx4 matrix containing the calculated parameters.
ECOFF	1x2 matrix containing the group index and ECOFF.
AIC	AIC value.
BIC	BIC value.
log_likelihood	1xM matrix containing the log likelihood value for each EM iteration. (M number of EM iterations)
EM_iterations	Number of EM iterations.

Author(s)

Lisa Allmesberger Fabian Bergs Stefan Immler Michael Kässmann

See Also

total.function

sp2019-antibiotics/Team-Student documentation built on Nov. 5, 2019, 9:13 a.m.