generate_data: Generate survival data for analysis

In sachsmc/pseupersims: Simulation Study for Predicting Pseudo-observations in Survival Analysis with Superlearner

Description

500 subjects, 20 variables, about 20 some linear, some wonky associations, an interaction, and some variables correlated.

Usage

generate_data(n = 500, scenario = "A", missing.p = 0.2)

Arguments

n	Sample size
scenario	Character designating the simulation scenario: 0 = NULL, no association, A = simple, one variable is linearly associated with IRIS, B = simple multivariate, 5 variables are linearly associated, C = wonky, interactions, nonlinearities, etc. D = same as B but censoring depends on covariates
missing.p	Proportion of missing binary Y values (random censoring)

Value

A data frame with X variables, censored survival times (competing risk w death), and true cumulative incidence at 26.5 weeks

sachsmc/pseupersims documentation built on June 12, 2019, 1:14 p.m.