ZINB_two_gamma: Zero-Inflated Negative Binomial - Two Gamma Component

In sidiwang/hgzips: hgzips

Description

This ZINB_two_gamma function finds hyperparameter estimates by implementing the Expectation-Maximization (EM) algorithm and zero-inflated negative binomial model with two gamma components. nlminb function is used to maximize the loglikelihood function.

Usage

ZINB_two_gamma(
  alpha1,
  beta1,
  alpha2,
  beta2,
  pi,
  omega,
  N,
  E,
  weight,
  iteration,
  Loglik = FALSE,
  zeroes = FALSE,
  N_star = 1
)

Arguments

alpha1	initial shape parameter value of the first gamma distribution for implementing the EM algprithm
beta1	initial rate parameter value of the first gamma distribution for implementing the EM algprithm
alpha2	initial shape parameter value of the second gamma distribution for implementing the EM algprithm
beta2	initial rate parameter value of the second gamma distribution for implementing the EM algprithm
pi	initial mixing proportion guess of the two gamma distributions
omega	initial weight for observing a true zero (according to zero-inflated poission distribution)
N	vector of Nij values
E	vector of Eij values
weight	set weight = rep(1, length(N)) if N and E are not squashed data, or input the weight vector corresponding to the squashed Nij vector.
iteration	number of EM algorithm iterations to run
Loglik	whether to return the loglikelihood of each iteration or not (TRUE or FALSE)
zeroes	A logical scalar specifying if zero counts should be included.
N_star	the minimum Nij count size to be used for hyperparameter estimation. If zeroes are included in Nij vector, please set N_star = NULL

Value

a list of estimated parameters and their corresponding loglikelihood if Nij = 0 is included in the input dataset

• theta_EM Estimate of hyperparameters for each EM iteration

• llh logliklihood for each EM iteration (optional)

if the minimum Nij count size included in the input dataset is not 0"

• theta_EM Estimate of hyperparameters for each EM iteration

See Also

openEBGM, nlminb

Examples

Nij = rnbinom(100, size = 2, prob = 0.3)
Eij = runif(100, 0, 2)
par_estimated = ZINB_two_gamma(0.2, 0.1, 2, 4, 0.33, 0.3, N = squashed$N, E = squashed$E, weight = squashed$weight, 10, Loglik = TRUE, zeroes = FALSE, N_star = 1)
par_estimated$theta_EM

sidiwang/hgzips documentation built on Jan. 19, 2021, 4:09 p.m.