关键词:
Data set generator
Survival data
Biomarker
Java
摘要:
Background: Software testing is an essential part of the software development process, but real-world data may not be suited or available for testing purposes. In the medical context, it can be especially hard to get the necessary test data for various reasons such as privacy concerns. To overcome these obstacles and provide data for the necessary thorough tests of software, the generation of simulated data sets can be a solution. In this paper, we focus on the challenging task of generating such survival data sets containing known effects. So far, no user-friendly software exists for the simulation of survival data, as they are typically derived from clinical trials with follow-ups. Results: To overcome these shortcomings, we developed an easy to use software package called vivaGen. In our Java software, parameters of survival time distributions are replaced by comprehensive measures that can be configured more intuitive by practitioners. vivaGen is equipped with a graphical frontend that allows users to adjust parameters and visualize the results in survival plots of the simulated cohorts. Conclusions: vivaGen is freely available and published as open source. It provides a novel way to generate test data sets based on probability distributions in a comprehensive and user-friendly way.