Nimble User Manual
User Manual:
Open the PDF directly: View PDF
Page Count: 192 [warning: Documents this large are best viewed by clicking the View PDF Link!]
- I Introduction
- II Models in NIMBLE
- III Algorithms in NIMBLE
- MCMC
- One-line invocation of MCMC: nimbleMCMC
- The MCMC configuration
- Building and compiling the MCMC
- User-friendly execution of MCMC algorithms: runMCMC
- Running the MCMC
- Extracting MCMC samples
- Calculating WAIC
- k-fold cross-validation
- Samplers provided with NIMBLE
- Detailed MCMC example: litters
- Comparing different MCMCs with MCMCsuite and compareMCMCs
- Sequential Monte Carlo and MCEM
- Spatial models
- Bayesian nonparametric models
- MCMC
- IV Programming with NIMBLE
- Overview
- Writing simple nimbleFunctions
- Creating user-defined BUGS distributions and functions
- Working with NIMBLE models
- Data structures in NIMBLE
- Writing nimbleFunctions to interact with models
- Overview
- Using and compiling nimbleFunctions
- Writing setup code
- Writing run code
- Driving models: calculate, calculateDiff, simulate, getLogProb
- Getting and setting variable and node values
- Getting parameter values and node bounds
- Using modelValues objects
- Using model variables and modelValues in expressions
- Including other methods in a nimbleFunction
- Using other nimbleFunctions
- Virtual nimbleFunctions and nimbleFunctionLists
- Character objects
- User-defined data structures
- Example: writing user-defined samplers to extend NIMBLE's MCMC engine
- Copying nimbleFunctions (and NIMBLE models)
- Debugging nimbleFunctions
- Timing nimbleFunctions with run.time
- Clearing and unloading compiled objects
- Reducing memory usage