run.Estep {coarseDataTools}R Documentation

The expectation step for a specialized EM algorithm

Description

The Expectation- or E-step for the EM algorithm implemented in EMforCFR(). Not meant to be called by users.

Usage

run.Estep(alpha, full.data, nlag, assumed.nu)

Arguments

alpha Current estimates of the alpha parameters from the GLM model.
full.data A matrix of observed data. See description in EMforCFR helpfile.
nlag The number of time units for lagged data. Corresponds to length(assumed.nu).
assumed.nu a vector of probabilities corresponding to the survival distribution, i.e. nu[i]=Pr(surviving i days | fatal case)

Value

A data matrix with the same format as full.data from the EMforCFR() documentation.

Author(s)

Nicholas G. Reich

See Also

EMforCFR


[Package coarseDataTools version 0.1 Index]