Related Experiment Videos
GENCOV: a Fortran program that generates randomly censored survival data with covariates.
A Ciampi1, J A Chapman, S A Hogg
1Montreal Children's Hospital Research Institute, Quebec, Canada.
Computer Methods and Programs in Biomedicine
|May 1, 1989
Summary
This study introduces a Fortran program for simulating censored survival data. It generates data for various distributions and regression models, aiding statistical research.
Area of Science:
- Biostatistics
- Computational Statistics
- Survival Analysis
Background:
- Censored survival data is common in medical research.
- Accurate simulation of this data is crucial for validating statistical methods.
- Existing simulation tools may lack flexibility in modeling complex survival patterns.
Purpose of the Study:
- To develop a versatile Fortran program for simulating censored survival data.
- To incorporate multiple parametric distributions and common regression models.
- To facilitate the generation of realistic survival data for research and validation.
Main Methods:
- The program utilizes Fortran for high-performance computation.
- It simulates survival times based on uniform, generalized Gamma, log-normal, and Pettitt's generalized logistic distributions.
- Box-Cox transformation and covariate inclusion enable flexible model specification.
Main Results:
- The program successfully generates censored survival data.
- It supports parametric Accelerated Failure Time, Proportional Hazards, and Proportional Odds regression models.
- Covariates can be integrated to create generalized log-gamma, log-normal, and Pettitt's regression models.
Conclusions:
- The developed Fortran program offers a robust tool for simulating censored survival data.
- It enhances the ability to study survival data under various parametric assumptions and regression frameworks.
- This simulation capability supports the advancement of statistical methodologies in survival analysis.