generate_prediction_intervals_1.2: Second tier function for generating prediction intervals from...

In jameshay218/mcmcJH: Generic random walk MCMC algorithm for parameter inference

Description

This function is called by a wrapper function, generate_prediction_intervals_1.1 to generate plottable prediction intervals for a model plot.

Usage

generate_prediction_intervals_1.2(mcmc_chains, burnin = NULL, times,
  runs = 100, level = 0.95, smoothing = 0, MODEL_FUNCTION)

Arguments

mcmc_chains	a matrix containing the whole MCMC chain (ie. parameter values at each iteration)
burnin	optional parameter specifying the number of iterations from the MCMC chain that should be discarded as burn in
times	a vector of time points to calculate the model trajectory over
runs	number of random draws from the multivariate posterior with which to calculate the prediction intervals. Defaults to 100
level	desired prediction interval size. Defaults to 0.95 ie. 95 percent prediction intervals
smoothing	uses smoothing splines to give smoother prediction intervals where these might be aethsetically spikey. Defaults to 0
MODEL_FUNCTION	a pointer to an existing R function that is used to calculate the model trajectories

Value

a list of the lower and upper prediction lines

See Also

generate_prediction_intervals_1.1

jameshay218/mcmcJH documentation built on May 18, 2019, 11:20 a.m.