is.magioutput: MagiSolver output ('magioutput') object

In magi: MAnifold-Constrained Gaussian Process Inference

MagiSolver output (magioutput) object

Description

Check for and create a magioutput object

Usage

is.magioutput(object)

magioutput(...)

Arguments

object	an R object
...	arguments required to create a magioutput object. See details.

Details

Using the core MagiSolver function returns a magioutput object as output, which is a list that contains the following elements:

theta

matrix of MCMC samples for the system parameters \theta, after burn-in.

xsampled

array of MCMC samples for the system trajectories at each discretization time point, after burn-in.

sigma

matrix of MCMC samples for the observation noise SDs \sigma, after burn-in.

phi

matrix of estimated GP hyper-parameters, one column for each system component.

lp

vector of log-posterior values at each MCMC iteration, after burn-in.

y, tvec, odeModel

from the inputs to MagiSolver.

Printing a magioutput object displays a brief summary of the settings used for the MagiSolver run. The summary method for a magioutput object prints a table of parameter estimates, see summary.magioutput for more details. Plotting a magioutput object by default shows the inferred trajectories for each component, see plot.magioutput for more details.

Value

logical. Is the input a magioutput object?

Examples

# Set up odeModel list for the Fitzhugh-Nagumo equations
fnmodel <- list(
  fOde = fnmodelODE,
  fOdeDx = fnmodelDx,
  fOdeDtheta = fnmodelDtheta,
  thetaLowerBound = c(0, 0, 0),
  thetaUpperBound = c(Inf, Inf, Inf)
)

# Example FN data
data(FNdat)

# Create magioutput from a short MagiSolver run (demo only, more iterations needed for convergence)
result <- MagiSolver(FNdat, fnmodel, control = list(nstepsHmc = 5, niterHmc = 50)) 

is.magioutput(result)

magi documentation built on April 26, 2023, 1:12 a.m.