simulate_multinomial_data {multinomialLogitMix}

Synthetic data generator

Description

This function simulates data from mixture of multinomial logistic regression models.

Usage

simulate_multinomial_data(K, p, D, n, size = 20, prob = 0.025, betaTrue = NULL)

Arguments

K

Number of clusters.

p

Number of covariates, including constant.

D

Number of multinomial categories.

n

Number of data points to simulate.

size

Negative Binomial parameter (number of successes). Default: 20.

prob

Negative Binomial parameter (probability of success). Default: 0.025.

betaTrue

An array which contains the true values of the logit coefficients per cluster. Default: randomly generated values.

Value

count_data

matrix of data counts.

design_matrix

design matrix.

clustering

Ground-truth partition of the data.

Author(s)

Panagiotis Papastamoulis