simpoisson: Simulate data for poisson regression

Description Usage Arguments Author(s)

View source: R/simpoisson.R

Description

Simulate from piecewise constant exponential regression model with two binary variables X, Z whose distributions and effects may vary in time intervals.

Usage

1
2
simpoisson(N, changepoints, baseline, beta, gamma, delta = 0, comprisk,
  alpha, zeta, px, pz, pv = NULL, pxz = 0, const = "none")

Arguments

N

sample size.

changepoints

borders of time intervals.

baseline

baseline rate in intervals.

beta

effect of X (log hazard ratio).

gamma

effect of Z (log hazard ratio).

delta

effect of V (log hazard ratio).

comprisk

rate of competing risk event.

alpha

effect of X on competing risk (log hazard ratio).

zeta

effect of Z on competing risk (log hazard ratio).

px

probability of X in intervals.

pz

probability of Z in intervals.

pv

probability vector for categories of V.

pxz

effect of Z on X (log odds ratio).

const

Whether the distributions of X and Z vary over time is controlled with argument ‘const’. const="XZ" means that both X and Z do not change over time. const="X" means that only X does not change over time. const="Z" means that only Z does not change over time.

Author(s)

Thomas A. Gerds <[email protected]>


tagteam/heaven documentation built on Feb. 16, 2019, 8:21 p.m.