Easy.Calibration: Easy.Calibration

In antonio-pgarcia/RRepast: Invoke 'Repast Simphony' Simulation Models

Description

Search for the best set of parameters trying to minimize the calibration function provided by the user. The function has to operational models, the first based on the experimental setup where all parameters are defined a priori and the second using optimization techniques. Currently the only supported optimization technique is the particle swarm optimization.

Usage

Easy.Calibration(m.dir, m.ds, m.time = 300, parameters, exp.n = 100,
  exp.r = 1, smax = 4, design = "lhs", FUN, default = NULL)

Arguments

m.dir	The installation directory of some repast model
m.ds	The name of any model aggregate dataset
m.time	The total simulated time
parameters	The input factors
exp.n	The experiment sample size
exp.r	The number of experiment replications
smax	The number of solutions to be generated
design	The sampling scheme ["lhs"|"mcs"|"ffs"]
FUN	The objective or cost function. A function defined over the model output.
default	The alternative values for parameters which should be kept fixed

Value

A list with holding experiment, object and charts

Examples

## Not run: 
 my.cost<- function(params, results) {
   criteria<- c()
   Rate<- AoE.RMSD(results$X.Simulated,results$X.Experimental)
   G<- AoE.RMSD(results$G.T.,52)
   total<- Rate + G
   criteria<- cbind(total,Rate,G)
   return(criteria)
 }
 
 Easy.Setup("/models/BactoSim")
 v<- Easy.Calibration("/models/BactoSim","ds::Output",360,
                       f,exp.n = 1000, exp.r=1, smax=4, 
                       design="mcs", my.cost)
 

## End(Not run)

antonio-pgarcia/RRepast documentation built on Feb. 22, 2020, 1:20 a.m.