SimTools: Output Analysis for Monte Carlo
- Mentors
- Dootika Vats, James Flegal
- Organization
- R project for statistical computing
- Technologies
- c++, r, latex, Rcpp, Roxygen2
- Topics
- visualization, simulation, statistics, Markov Chain Monte Carlo (MCMC), Output Analysis, Confidence Interval
I propose completing the R package "SimTools" before it's ready for CRAN submissions. To equip the user with almost all types of output analysis of MCMC, SimTools is here. SimTools is an R package (not completed) that provides users with an enriched analysis of MCMC simulations (also iids) and facilitates the generation of Markov Chains. This project shall affect both beginners and well-experienced users. More examples of how to write complicated simulation and analysis methods will help beginners, and they will be able to use the models in their projects directly. Importing simulated chains from other libraries will help the experienced users perform transfer learning on their Simulated Markov Chains. This tool provides a range of MCMC algorithms and diagnostic tools to help users evaluate the performance of their models. As the Package is Partially developed, I need to implement six significant tasks to complete this package(Reference - Project webpage). These tasks are Improving Multiple chain compatibility, Implementing Summary function for output, Compatibility for discrete state space Markov Chains, Efficient trace plots, Export to C++(CI Calculations), and Rich Documentation of the Package. I have already gone through all these problems and developed possible approaches to solve them.