RunDesign: Call an external (to R) code wrapped through Funz...

In Funz/Funz.R: Parametric Computing Environment

Call an external (to R) code wrapped through Funz environment.

Description

Call an external (to R) code wrapped through Funz environment.

Usage

RunDesign(
  model = NULL,
  input.files,
  input.variables = NULL,
  output.expressions = NULL,
  design = NULL,
  design.options = NULL,
  run.control = list(force.retry = 2, cache.dir = NULL),
  monitor.control = list(results.tmp = TRUE, sleep = 5, display.fun = NULL),
  archive.dir = NULL,
  out.filter = NULL,
  verbosity = 1
)

Arguments

model	name of the code wrapper to use. See .Funz.Models global var for a list of possible values.
input.files	list of files to give as input for the code.
input.variables	list of variables definition in a String (for instance x1="[-1,1]"), or array of fixed values (will launch a design for each combination).#' @param all.combinations if FALSE, input.variables variables are grouped (default), else, if TRUE experiments are an expaanded grid of input.variables
output.expressions	list of interest output from the code. Will become the names() of return list.
design	Design of Experiments (DoE) given by its name (for instance ""). See .Funz.Designs global var for a list of possible values.
design.options	list of options to pass to the DoE. All options not given are set to their default values. Note that '_' char in names will be replaced by ' '.
run.control	list of control parameters: 'force.retry' is number of retries before failure, 'cache.dir' setup array of directories to search inside before real launching calculations.
monitor.control	list of monitor parameters: sleep (delay time between two checks of results), display.fun (function to display project cases status).
archive.dir	define an arbitrary output directory where results (cases, csv files) are stored.
out.filter	what output(s) to retreive in returned object.
verbosity	0-10, print information while running.

Value

list of array design and results from the code.

Examples

## Not run: 
# Search for minimum of Branin function, taken as the model (test case)
install.Design("GradientDescent") # Download on github the GradientDescent algorithm
startCalculators(5) # start calculator instances to run model
RunDesign(model="R",
          input.files=file.path(Funz:::FUNZ_HOME,"samples","branin.R"),
          output.expressions="cat", design = "GradientDescent",
          design.options = list(max_iterations=10),input.variables = list(x1="[0,1]",x2="[0,1]"))

# More realistic case using inversion of Modelica:
#  find convection coefficient that gives minimal temperature of 40 degrees.
install.Model("Modelica") # Install Modelica plugin (Assumes that Modelica is already installed)
install.Design("Brent")   # Install Brent algorithm for inversion
calcs = startCalculators(5) # Will start calculator instances, later used by Run()
NewtonCooling = RunDesign(model = "Modelica",
  input.files = file.path(Funz:::FUNZ_HOME,"samples","NewtonCooling.mo.par"),
  input.variables = list(convection="[0.0001,1]"), output.expressions = "min(T)",
  design="Brent",design.options=list(ytarget=40.0))
plot(NewtonCooling$convection[[1]], NewtonCooling$`min(T)`[[1]])
abline(h=40.0)
abline(v=NewtonCooling$analysis.root)
stopCalculators(calcs) # Shutdown calculators (otherwise stay in background)

## End(Not run)

Funz/Funz.R documentation built on Jan. 28, 2024, 11:39 p.m.