icsurv {Icens}R Documentation

The class of objects returned by the estimation routines in the Icens library.

Description

An object of class icsurv must contain the following components:

converge
A boolean indicating whether the iteration producing pf converged.
pf
The probability vector.

It can optionally contain any of the following components:

clmat
The clique matrix used to obtain pf.
intmap
The real representations of the support for the components of pf.
iter
A matrix containing every iterative estimate of pf, useful for debugging.
lval
The value of the log likelihood at pf.
numiter
The number of iterations taken.
sigma
The cumulative sum of pf.
weights
Weights used in the EMICM algorithm.

Author(s)

Alain Vandal and Robert Gentleman.

See Also

VEM, ISDM, EMICM, PGM, EM


[Package Icens version 0.5 Index]