fit_stan_models: Fit and summarise models

In seabbs/DirectEffBCGPolicyChange: Estimating the direct effect of the 2005 UK BCG vaccination policy change.

Description

Fit and summarise models

Usage

fit_stan_models(ModelReadyDataSpec, PreCompile = TRUE,
  BaseLineModelsNames = NULL, FileName = "ModelResults", SubFiles = 4,
  Folder = "results", Path = ".", ReRunAnalysis = FALSE, verbose = TRUE,
  Cores = NULL, Iterations = 10000, BurnIn = 5000, Alpha = 0.99,
  Treedepth = 10, UpdateAlpha = 0.99999, HandleDLLs = FALSE,
  UpdateTreedepth = 20)

Arguments

ModelReadyDataSpec	A dataframe of data ready for modelling
PreCompile	Logical, defaults to TRUE. Should a subset of models be precompiled (specified with BaseLineModelNames) for the first vaccination scheme and population?
BaseLineModelsNames	A character string of baseline model names.
FileName	A character string to use as the file name for the results. Defaults to ModelResults.
SubFiles	Numeric, defaults to 4. The number of subfiles to split the results into. Results are saved as ModelResults1 etc. This must be smaller or equal to the number of models to be fitted.
Folder	Character string to use as the folder name for the results. Defaults to "results"
Path	Character string specifying the path to the results folder. Defaults to ".".
ReRunAnalysis	Logical, default to FALSE. Should the analysis be rerun.
verbose	Logical defaults to FALSE. Should progress messages be printed.
Cores	Numeric, defaults to the number of cores available minus 1. How many cores to use for the analysis.
Iterations	Numeric, the number of iterations to use for the MCMC sampler. Defaults to 10000.
BurnIn	Numeric, the number of iterations to use for the burn in. Defaults to 5000.
Alpha	Numeric, set the adaptive alpha value for the MCMC. Defaults to 0.99
Treedepth	Numeric, set the maximum tree depth. Defaults to 10.
UpdateAlpha	Numeric, set an updated alpha for models that have failed to fit. Defaults to 0.99999.
HandleDLLs	Logical, defaults to FALSE. Enables tracking and unloading of Stan DLL objects. If not reachng R's DLL cap then this is not required.
UpdateTreedepth	Numeric, set the updated maximum tree depth for models that have failed to fit.

Value

A dataframe of summarised results for the models specified to be fitted in the input dataframe.

See Also

clear_user_DLL Base_Model summarise_model_fit

seabbs/DirectEffBCGPolicyChange documentation built on Dec. 14, 2019, 9:24 p.m.