importance_sampling: Calculate Importance Weights

In codatmo/stanIncrementalImportanceSampling: Provides Functionality for Calculating Importance Weights for Stan MCMC Samples

Description

This function calculates importance weights for a previously fitted Stan model given a new set of input data. Given a stanfit object with MCMC samples and a new set of input data for the stan model, calculate importance sampling weights. For example, a time series model predicting into the future can be updated with new data post-hoc when data becomes available.

Usage

importance_sampling(
  originalFit,
  oldData,
  newData,
  model_code = originalFit@stanmodel@model_code[1]
)

Arguments

originalFit	stanfit containing MCMC samples
oldData	list() containing data used to fit originalFit
newData	list() containing the new input data
model_code	string containing Stan model code for originalFit (optional)

Value

list containing weights, neff, lpOriginal and lpNew

codatmo/stanIncrementalImportanceSampling documentation built on Dec. 19, 2021, 5:21 p.m.