The Respondent Driven Sampling (RDS) Simulator is designed to aid researchers with basic questions of sample design.
For latest version, please check out by subversion.
The Respondent Driven Sampling (RDS) Simulator is designed to aid researchers with basic questions of sample design
• What is the optimal number of seeds?
• What is the optimal number of coupons to give each responent?
• What is the probability of the recruitment chain dying out?
• How long is will it take for the sample to be collected?
• What is the probability of the sample completing before the desired finish date?
Unfortunately, there are no conventional mathematics to answer these questions. RDS Sim. overcomes this hurdle by simulating numerous recruitment chains in continuous time. This allows you to investigate the ramifications of changing sample design on the timing and logistics of your recruitment chain.
There is no documentation, but RDS Sim. has been designed to be extremely easy to use. Contact the author with any questions, feature requests or bug-reports.
The first version (0.01) is now available in a Windows installer. If you use Mac or Linux (good for you), you can download the source-code and run it with a python interpreter. RDS Simulator is free, open-source, and licensed under the GPL.
The windows version has been tested on XP, but depends on several DLLs. If you have any problems due to failed dependencies, contact the author.
The source-code version is platform-independent, but requires python and depends on several libraries: Matplotlib and scipy. Future versions will likely require networkx.