Learn R Programming

⚠️There's a newer version (2.11.14) of this package.Take me there.

SPOT (version 2.11.10)

Sequential Parameter Optimization Toolbox

Description

A set of tools for model-based optimization and tuning of algorithms (hyperparameter tuning respectively hyperparameter optimization). It includes surrogate models, optimizers, and design of experiment approaches. The main interface is spot, which uses sequentially updated surrogate models for the purpose of efficient optimization. The main goal is to ease the burden of objective function evaluations, when a single evaluation requires a significant amount of resources.

Copy Link

Version

Install

install.packages('SPOT')

Monthly Downloads

324

Version

2.11.10

License

GPL (>= 2)

Maintainer

Thomas Bartz-Beielstein

Last Published

May 27th, 2022

Functions in SPOT (2.11.10)

Build Kriging Model

buildGaussianProcess

Gaussian Process Model Interface

Build LOESS Model

Sequential Parameter Optimization Toolbox

Linear Model Interface

buildEnsembleStack

Ensemble: Stacking

Bayesian Optimization Model Interface

buildKrigingDACE

Build DACE model

ranger Interface

calculationMothes

Cyclone Simulation: Mothes

buildRandomForest

Random Forest Interface

checkFunEvalsDesignSize

Check funEvals Setting against designSize

Check for NAs in x lower and upper

Build Response Surface Model

checkFeasibilityNlopGnIngres

Check feasibility for NLOPT_GN_ISRES

Build random search dummy model

calculationBarthMuschelknautz

Cyclone Simulation: Barth/Muschelknautz

checkTypesOfInput

Check Input Types

Correlation: Exp

checkVerbosityLevels

Check correct verbosity levels

Correlation: Expg

DACE objective function

Wrapper for Maximum Likelihood Estimation

Descent RSM model

Correlation: Spherical

checkInputDimensionsionalityCorrect

Check Dimensions of spotInputs

Gas Sensor Data

corrnoisykriging

Correlation: Noisy Kriging

expectedImprovement

Expected Improvement

Lasso Model Interface

checkInputTypesInControl

Check input types in the spotControl list.

checkLowerSmallerThanUpper

#' Check That Lower and Upper are not Equal #' #' If any entries in lower and upper are equal, the parameter has no range and cant be optimized. #' In that case, an error is generated. #' #' @param lower is a vector that defines the lower boundary of search space. This determines also the dimensionality of the problem. #' @param upper is a vector that defines the upper boundary of search space. #' #' @keywords internal checkLowerNotEqualsUpper <- function(lower,upper) if(any(lower == upper)) stop("SPOT Configuration Error: Entries in 'lower' and 'upper' should not be equal") Check That Lower is smaller than Upper

Prepare DACE fit

Correlation: Kriging

Fix model parameters DACE

Correlation: Gauss

Correlation: Noisy Gauss

Normalized LHD Design

Latin Hypercube Design Generator

designUniformRandom

Uniform Design Generator

funOptimLecture

funOptimLecture

Correlation: Lin

Transform coded values to natural values

Correlation: Cubic

Ishigami Test Function (No. 7)

funRosen (No. 2, More No. 1)

funPowellS (No. 19, More No. 13)

funHelical (No. 13, More No. 7)

Sobol and Levitan Test Function (No. 6)

Evaluate DACE fit

getPerformanceStats

get performance stats

funPowellBs (No. 9, More No. 3)

getNatDesignFromCoded

Get natural parameter values from coded +-1 representation

Correlation: Spline

daceStartParameters

Start parameter setup DACE

funBard (No. 14, More No. 8)

Parallel execution of code, dependent on the operating system

funShiftedSphere

funShiftedSphere (No. 4)

funGauss (No. 15, More No. 9)

handleNAsKrigingWorst

handleNAsKrigingWorst

funFreudRoth (No. 8, More No. 2)

funRosen2 (No. 2a)

duplicateAndReplicateHandling

duplicateAndReplicateHandling

funbrownBs (No. 10, More No. 4)

funSphere (No. 3)

makeSpotFunList

makeSpotFunList

infillGetFullPrediction

infillGetFullPrediction

get number of replicates

Evolution Strategy

funGulf (No.17, More No. 11)

funBeale (No.11, More No. 5)

funGoldsteinPrice

Goldstein-Price Test Function (No. 5)

funBranin (No. 1)

funbox3D (No. 18, More No. 12)

maxNearestNeighbourDistance

maxNearestNeighbourDistance

funJennSamp (No. 12, More No 6)

Random search surrogate-optimizer

prepareBestObjectiveVal

Preprocess y Values to Plot Best Objective Value

Plot Best Objective Value

predict.spotBOModel

Prediction method for bayesian optimization model

Minimization by GENetic Optimization Using Derivatives

Interpolated plot

predict.spotGaussianProcessModel

Prediction method for Gaussian Process Model

funMeyer (No. 16, More No. 10)

get impute positions

Objective function - Cyclone Simulation: Barth/Muschelknautz

initialInputCheck

initialInputCheck. Initial Input Check of Spot Configuration

Impute NAs and Inf in y

getMultiStartPoints

Get Multi Start Points

makeMoreFunList

makeMoreFunList

linearAdaptedSE

linearAdaptedSE

Wrapper for Expected improvement (Gramacy)

getCorrelationMatrix

getCorrelationMatrix

Print Function DACE Kriging

plotPCAvariance

plotPCAvariance

S-Ring Simulation Data Obtained With SPOT

predict.ensembleStack

Predict Stacked Ensemble

print.spotGaussianProcessModel

Print method for Gaussian Process Model

print.spotBOModel

Print method for BO model

objectiveFunctionEvaluation

objectiveFunctionEvaluation Objective Function Evaluation

simulateFunction

simulateFunction

simulate.kriging

Kriging Simulation

Expected Improvement Infill Criterion

Minimization by Differential Evolution

normalizeMatrix

Normalize design matrix

normalizeMatrix2

Normalize design 2

infillExpectedImprovement

infillExpectedImprovement

Print method for RSM model

predict.spotLOESS

Predict loess model

print.spotLinearModel

Print method for linear model

krigingLikelihood

Calculate negative log-likelihood

Regression: Regpoly2

print.spotRanger

Print method for random forest

predict.spotLassoModel

Prediction method for lasso model

print.spotRandomForest

Print method for random forest

result from the vignette benchmark

S-Ring Simulation Data Obtained With SPOT

predict.cvModel

predict.cvModel

plot.spotTreeModel

Plot rpart tree model

plot.spotSeverity

Plot method for spotSeverity

plotSingleDimFunction

plotSingleDimFunction

Minimization by L-BFGS-B

optimNLOPTR. Minimization by NLOPT

predict.spotRanger

Predictor for spotExtraTrees

Surface plot of a model

Interface to minimization based on Gramacy's lagp package

Surface plot of a function

predict.kriging

Predict Kriging Model

predict.spotRandomForest

Prediction method for random forest

predict.rsdummy

Predict random search dummy

Interface SANN to SPOT

repairNonNumeric

Repair Non-numeric Values

Regression: Regpoly0

Satterthwaite Function

spotAlgEsDominantReco

spotAlgEsDominantReco

spotAlgEsGetSuccessRate

get Success Rate

selectN. Select n Design Points

S-Ring Simulation Data

Termination hps

predictKrigingReinterpolation

Predict Kriging Model (Re-interpolating)

spotAlgEsIndividualInitial

Individual Initialization

Minimization by Latin Hypercube Sampling

Expected improvement (Gramacy)

predict.spotTreeModel

Prediction method for rpart tree models

Regression: Regpoly1

print.spotSeverity

Print method for spotSeverity

spotAlgEsMarriage

spotAlgEsInterRecoBeSw02

spotAlgEsInterRecoBeSw02

spotAlgEsStratMutation

spotAlgEsStratMutation

spotAlgEsSelection

spotAlgEsSelection

print.spotTreeModel

Print method for rpart tree models

spotFillControlList

spotFillControlList

predict.spotLinearModel

Prediction method for linear model

print.spotLOESS

Print method for loess model

spotSeverityBasic

spotSeverityBasic

predict.spotRSM

Predict RSM model

Function Evaluation Wrapper

simulationSpectral

simulationSpectral

simulationDecompose

Kriging Simulation: Decomposition

wrapFunctionParallel

Parallelized Function Evaluation Wrapper

wrapSystemCommand

wrapSystemCommand

spotPlotSeverityBasic

spotPlotSeverityBasic

Optimal Computing Budget Allocation

print.spotLassoModel

Print method for lasso model

Sequential Parameter Optimization Main Loop

spotAlgEsInitParentPop

Initialize Parent Population

Backslash operator.

spotAlgEsInterReco

spotAlgEsInterReco

Evolution Strategy Implementation

spotAlgEsMarriageWithReplace

Marriage with replace

S-Ring Simulation Data

thetaNuggetGradient

thetaNuggetGradient

Transform input

S-Ring Simulation Data

spotAlgEsObjMutation

spotAlgEsObjMutation

spotAlgEsTermination

formatted output dependent on verbosity

updateSeedFromXnew

wrapSystem_parseMatrixToString

wrapSystem_parseMatrixToString