score: Gradient of the log-likelihood

In wrbrooks/latent: Estimate the parameters of a Poisson latent variable model with left-censoring

Description

Evaluates the score functions (gradients of the log-likelihood) for a Poisson distribution with log-linear model for the mean.

Usage

score(data, params, event, specific = NULL)

Arguments

data	matrix of observations, with rows equal to the number of observations and columns equal to the number of categories
params	vector of model parameters, consisting of \boldsymbol{α}, \boldsymbol{β}, \boldsymbol{γ}.
event	vector of labels with one entry per row of data, where each entry identifies the event to which the corresponding row belongs
specific	vector of logical values, one entry per column of data, with each entry indicating whether the corresponding column of the data is a human-specific bacterium

Value

The score functions, evaluated at the current parameter values.

wrbrooks/latent documentation built on May 4, 2019, 11:59 a.m.