runoptDirect: runoptDirect Optimierung der babsim.hospital Parameter

Description

SPOT Aufruf zur Optimierung der babsim Parameter mit Lasso

Usage

runoptDirect(
  expName = "obkpara20201017",
  rkiwerte = babsim.hospital::rkidata,
  icuwerte = babsim.hospital::icudata,
  region = 5374,
  TrainFieldStartDate = NULL,
  TrainSimStartDate = NULL,
  TestFieldStartDate = NULL,
  TestSimStartDate = NULL,
  Overlap = 7,
  verbosity = 0,
  seed = 123,
  direct = FALSE,
  repeats = 1,
  funEvals = 35,
  funEvalsFactor = 0,
  size = 30,
  simrepeats = 2,
  subset = 32,
  parallel = FALSE,
  percCores = NULL,
  icu = TRUE,
  icuWeights = 1,
  testRepeats = 3,
  resourceNames = c("intensiveBed", "intensiveBedVentilation"),
  resourceEval = c("intensiveBed", "intensiveBedVentilation"),
  spotEvalsParallel = FALSE,
  tryOnTestSet = TRUE
)

Arguments

expName

Experiment Name

rkiwerte

RKI Daten

icuwerte

ICU Daten

region

Landkreis Id, e.g., 5374 fuer OBK, 5315 fuer Koeln, 0 fuer Deutschland, oder Bundesland ID, e.g., 5 fuer NRW.

TrainFieldStartDate

Start (Tag), e.g., "2020-06-01"

TrainSimStartDate

Start (Tag), e.g., "2020-05-01"

TestFieldStartDate

Start (Day), e.g., "2020-06-01" for test field data

TestSimStartDate

Start (Day), e.g., "2020-05-01" for test simulation data, TestSimStartDate is usually before TestFieldStartDate

Overlap

integer. Days, train data will be extended (overlap with test data). Default: 7

verbosity

verbosity (int). Default: 0

seed

Seed

direct

use model-free optimization. Default: FALSE

repeats

Wiederholungen fuer SPOT (Optimierungslaeufe mit unterschiedlichem Seed)

funEvals

Auswertungen fuer SPOT (Simulationen, die fuer einen SPOT Lauf zur Verfuegung stehen)

funEvalsFactor

factor to increase function evaluations. Default: 0

size

Groesse des initialen Desings

simrepeats

Sim Wdhlg

subset

Subset (SPOT)

parallel

logical

percCores

percentage

icu

ICU Daten

icuWeights

Gewichtung der ICU Betten

testRepeats

number of final evaluations on the test data

resourceNames

Name der Ressourcen

resourceEval

Name der zu evaluierenden Ressourcen

spotEvalsParallel

Should the spot repeats be evaluated in parallel?

tryOnTestSet

Should results be tested on a separate test set?