Gaussian log-density with optional left censoring at an LOD

Detected observations contribute a Gaussian log-density on the measurement scale. Left-censored observations contribute the Gaussian probability of falling at or below the LOD.

Usage

mbmm_dnorm_censored(y, mean, sd, censored = FALSE, lod = NULL)

Arguments

y

Observed value on the same scale as mean. Values of y are ignored where censored is TRUE.

mean

Predicted mean on the likelihood scale.

sd

Strictly positive measurement-error standard deviation.

censored

Logical or 0/1 vector indicating left-censored observations.

lod

LOD on the same scale as y and mean. Required for every censored observation.

Value

Numeric vector of pointwise log-likelihood contributions.

Details

All arguments must already be on the same likelihood scale, normally the log10 scale used by MBMM.