Difference between revisions of "Config:SequentialDesign"

From SUMOwiki
Jump to navigationJump to search
 
(4 intermediate revisions by 2 users not shown)
Line 1: Line 1:
'''Generated for SUMO toolbox version 6'''.
+
'''Generated for SUMO toolbox version 7.0'''.
''We are well aware that the list below is incomplete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower.  The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files.''
+
''We are well aware that documentation is not always complete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower.  We are are a university research group after all. The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files.  If something is unclear please dont hesitate to [[Reporting problems|ask]].''
 
== SampleSelector ==
 
== SampleSelector ==
  
 
=== empty ===
 
=== empty ===
Dont select any new samples, useful when modeling multiple outputs in paralel
+
Don't select any new samples, useful when modeling multiple outputs, and you don't want to involve one of these outputs in the sampling process.
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#EmptySampleSelector|EmptySampleSelector]]" combineOutputs="false"/>
 
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#EmptySampleSelector|EmptySampleSelector]]" combineOutputs="false"/>
 
</source>
 
</source>
 
=== random ===
 
=== random ===
Each sampling iterations new samples are selected randomly
+
Selects new samples randomly in the design space.
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#RandomSampleSelector|RandomSampleSelector]]" combineOutputs="false"/>
 
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#RandomSampleSelector|RandomSampleSelector]]" combineOutputs="false"/>
 
</source>
 
</source>
=== combo ===
+
=== delaunay ===
Allows you combine multiple sample selector algorithms
+
This sample selector uses a Delaunay triangulation of the data to select samples in locations far from previous samples, or in locations where the estimated model error is largest. This algorithm uses QHull, which is very slow for high dimensions, so you should only use this sample selector for less than 6D and for less than 1000 samples.
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#CombinedSampleSelector|CombinedSampleSelector]]" combineOutputs="false">
+
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#PipelineSampleSelector|PipelineSampleSelector]]" combineOutputs="false">
   <[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#RationalPoleSuppressionSampleSelector|RationalPoleSuppressionSampleSelector]]" combineOutputs="false">
+
 
      <!-- Currently no options are available, if the model is a rational model, and a
+
   <[[Config:CandidateGenerator|CandidateGenerator]] type="[[CandidateGenerator#DelaunayCandidateGenerator|DelaunayCandidateGenerator]]"/>
      - - pole is found (using a simple search strategy), the pole is returned -->
 
  </[[Config:SampleSelector|SampleSelector]]>
 
 
    
 
    
  <[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#DelaunaySampleSelector|DelaunaySampleSelector]]" combineOutputs="false">
+
        <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#modelDifference|modelDifference]]">
      <!-- One of all, data -->
+
          <Option key="criterion_parameter" value="2"/>
      <Option key="sampleSelect" value="all"/>
+
        </[[Config:CandidateRanker|CandidateRanker]]>
      <!-- Integer between 2 and 20 -->
+
        <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#delaunayVolume|delaunayVolume]]"/>
      <Option key="nLastModels" value="2"/>
+
       
      <!-- One of densityBased, differenceBased, weightedLinear, weightedGeometric -->
+
        <[[Config:MergeCriterion|MergeCriterion]] type="[[MergeCriterion#WeightedAverage|WeightedAverage]]" weights="[1 1]"/>
      <Option key="scoreFunction" value="weightedLinear"/>
+
       
      <!-- Weighting for weightedLinear -->
 
      <Option key="lambda" value="0.5"/>
 
      <!-- Weighting for weightedGeometric -->
 
      <Option key="mu" value="0.5"/>
 
      <!-- One of none, max, cap, capmax -->
 
      <Option key="volumeScaling" value="max"/>
 
      <Option key="differenceScaling" value="capmax"/>
 
      <!-- Boolean flag, if set all points closer than snapThreshold to the boundary of
 
      - - the domain are clipped to the boundary -->
 
      <Option key="snapToEdge" value="enable"/>
 
      <Option key="snapThreshold" value=".05"/>
 
  </[[Config:SampleSelector|SampleSelector]]>
 
</[[Config:SampleSelector|SampleSelector]]>
 
</source>
 
=== delaunay ===
 
An adaptive sample selection algorithm that does a trade-off between error and density
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#DelaunaySampleSelector|DelaunaySampleSelector]]" combineOutputs="false">
 
  <!-- One of all, data -->
 
  <Option key="sampleSelect" value="all"/>
 
  <!-- Integer between 2 and 20 -->
 
  <Option key="nLastModels" value="2"/>     
 
  <!-- One of densityBased, differenceBased, weightedLinear, weightedGeometric -->
 
  <Option key="scoreFunction" value="weightedLinear"/>
 
  <Option key="lambda" value="0.5"/>
 
  <Option key="mu" value="0.5"/>
 
  <!-- One of none, max, cap, capmax -->
 
  <Option key="volumeScaling" value="max"/>
 
  <Option key="differenceScaling" value="capmax"/>
 
  <!-- Boolean flag -->
 
  <Option key="snapToEdge" value="enable"/>
 
  <Option key="snapThreshold" value=".2"/>
 
 
</[[Config:SampleSelector|SampleSelector]]>
 
</[[Config:SampleSelector|SampleSelector]]>
 
</source>
 
</source>
 
=== density ===
 
=== density ===
A simple density based sample selection algorithm
+
A space-filling sampling algorithm which uses an approximation of the Voronoi tessellation of the design space. Will only sample within the "allowed" areas if constraints are specified.
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#DensitySampleSelector|DensitySampleSelector]]" combineOutputs="false"/>
+
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#VoronoiSampleRanker|VoronoiSampleRanker]]" combineOutputs="false"/>
 
</source>
 
</source>
 
=== error ===
 
=== error ===
An adaptive sample selection algorithm (error based), driven by the evaluation of your model on a dense grid
+
An adaptive sample selection algorithm (error based), driven by the evaluation of your model on a dense grid, which selects samples in locations where the model error is estimated to be the largest.
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#ErrorSampleSelector|ErrorSampleSelector]]" combineOutputs="false">
+
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#PipelineSampleSelector|PipelineSampleSelector]]" combineOutputs="false">
  <!-- Integer between 2 and 20 -->
+
 
   <Option key="nLastModels" value="4"/>    
+
   <[[Config:CandidateGenerator|CandidateGenerator]] type="[[CandidateGenerator#GridCandidateGenerator|GridCandidateGenerator]]"/>
   <!-- One of none, max, cap, capmax -->
+
    
   <Option key="differenceScaling" value="none"/>
+
   <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#modelDifference|modelDifference]]">
  <!-- Gridsize to evaluate on, one of int or array of dimension length -->
+
          <Option key="criterion_parameter" value="4"/>
  <Option key="gridSize" value="50"/>
+
        </[[Config:CandidateRanker|CandidateRanker]]>
  <!-- Maximum total points to evaluate, distributed over dimensions -->
+
       
  <Option key="maxGridSize" value="100000"/>
+
        <[[Config:MergeCriterion|MergeCriterion]] type="[[MergeCriterion#ClosenessThreshold|ClosenessThreshold]]">
   <!-- Closeness threshold, Double -->
+
    
  <Option key="closenessThreshold" value="0.2"/>
+
      <!-- Closeness threshold, Double -->
  <!-- Set a % of the maximumSamples to randomly chosen -->
+
      <Option key="closenessThreshold" value="0.05"/>
  <Option key="randomPercentage" value="20"/>
+
      <!-- Set a % of the maximumSamples to randomly chosen -->
 +
      <Option key="randomPercentage" value="20"/>
 +
 
 +
      <Option key="debug" value="off"/>
 +
  </[[Config:MergeCriterion|MergeCriterion]]>
 
</[[Config:SampleSelector|SampleSelector]]>
 
</[[Config:SampleSelector|SampleSelector]]>
 
</source>
 
</source>
=== gradient ===
+
=== lola-voronoi ===
A highly adaptive sampling algorithm, error and density based
+
A highly adaptive sampling algorithm which performs a trade-off between exploration (filling up the design space as equally as possible) and exploitation (selecting data points in highly nonlinear regions). lola-voronoi is the only sample selector which currently supports multiple outputs, auto-sampled inputs and constraints.
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#GradientSampleSelector|GradientSampleSelector]]" combineOutputs="false">
+
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#LOLAVoronoiSampleSelector|LOLAVoronoiSampleSelector]]" combineOutputs="false">
 
   <!-- Integer between 2 and 20 -->
 
   <!-- Integer between 2 and 20 -->
 
   <Option key="neighbourhoodSize" value="2"/>
 
   <Option key="neighbourhoodSize" value="2"/>
 +
  <!-- Number of frequency values returned for each submitted sample. Only used with auto-sampled inputs. -->
 +
  <Option key="frequencies" value="11"/>
 
</[[Config:SampleSelector|SampleSelector]]>
 
</[[Config:SampleSelector|SampleSelector]]>
 
</source>
 
</source>
=== autoSampling ===
+
=== rationalPoleSupression ===
A wrapper around another sample selector that filters out the auto-sampled dimensions.
+
A sampling algorithm aimed at supressing poles in rational models by sampling them (only for Rational models)
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#AutoSamplingSampleSelector|AutoSamplingSampleSelector]]" combineOutputs="false">
+
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#OptimizeCriterion|OptimizeCriterion]]" combineOutputs="false">
 
+
 
   <!-- The filter function to use. Default = max. -->
+
   <!-- This criterion has to be solved to choose new samples, one can choose the optimizer used here -->
  <Option key="function" value="max"/>
+
   <[[Config:Optimizer|Optimizer]]>patternsearch</[[Config:Optimizer|Optimizer]]>
 
 
  <!-- The sample selector to use for non-auto sampled dimensions. -->
 
   <[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#GradientSampleSelector|GradientSampleSelector]]" combineOutputs="false">
 
      <Option key="neighbourhoodSize" value="2"/>
 
  </[[Config:SampleSelector|SampleSelector]]>
 
 
    
 
    
 +
  <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#rationalPoleSupression|rationalPoleSupression]]" scaling="none"/>
 +
  <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#modelDifference|modelDifference]]" scaling="none"/>
 +
 +
  <!--
 +
  when debug is 'on' a contour plot of the criterion function is drawn every iteration.
 +
  Together with the current samples and the chosen samples
 +
  -->
 +
  <Option key="debug" value="off"/>
 
</[[Config:SampleSelector|SampleSelector]]>
 
</[[Config:SampleSelector|SampleSelector]]>
 
</source>
 
</source>
=== isc ===
+
=== expectedImprovement ===
A sampling algorithm aimed at optimization problems
+
A sampling algorithm aimed at optimization problems (only for Kriging and RBF)
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#InfillSamplingCriterion|InfillSamplingCriterion]]" combineOutputs="false">
+
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#OptimizeCriterion|OptimizeCriterion]]" combineOutputs="false">
  <!-- A criterion determines the interesting regions to sample -->
 
  <!-- Choose 1 from the following: -->
 
  <Option key="criterion" value="gei"/> <!-- generalized expected improvement -->
 
  <Option key="g" value="1"/> <!-- balanced local-global search -->
 
  
   <!--<Option key="criterion" value="wei" />--> <!-- weighted expected improvement -->
+
   <!-- This criterion has to be solved to choose new samples, one can choose the optimizer used here -->
   <!--<Option key="w" value="0.5" />--> <!-- weight, 0 is global search, 1 is local search -->
+
   <[[Config:Optimizer|Optimizer]]>patternsearch</[[Config:Optimizer|Optimizer]]>
 
 
  <!--<Option key="criterion" value="ei" />--> <!-- expected improvement -->
 
 
    
 
    
   <!--<Option key="criterion" value="kushner" />--> <!-- kushner -->
+
   <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#expectedImprovement|expectedImprovement]]" scaling="none">
   <!--<Option key="eps" value="0.001" />-->
+
   </[[Config:CandidateRanker|CandidateRanker]]>
 
+
   <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#maxvar|maxvar]]" scaling="none"/>
  <!--<Option key="criterion" value="lcb" />--> <!-- lower confidence bound function -->
 
  <!--<Option key="lcb" value="2.0" />-->
 
  <!--<Option key="criterion" value="maxvar" />--> <!-- maximizes variation -->
 
 
 
  <!-- Watson and Barnes criterions -->
 
  <!--<Option key="criterion" value="wb1" />--> <!-- threshold-bounded extreme -->
 
  <!--<Option key="criterion" value="wb2" />--> <!-- regional extreme -->
 
 
 
  <!--<Option key="criterion" value="crowdedness" />--> <!-- crowdedness function -->
 
 
 
  <!-- This criterion has to be solved to choose new samples, one can choose the optimizer used here -->
 
   <[[Config:Optimizer|Optimizer]] type="[[Optimizer#DirectOptimizer|DirectOptimizer]]">
 
      <Option key="maxevals" value="1000"/>
 
      <Option key="maxits" value="300"/>
 
  </[[Config:Optimizer|Optimizer]]>
 
  
 
   <!--
 
   <!--
  <[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabGA|MatlabGA]]">
+
   when debug is 'on' a contour plot of the criterion function is drawn every iteration.
  </[[Config:Optimizer|Optimizer]]>
 
  -->
 
 
 
  <!--
 
   when debug is 'on' a contour plot of the ISC function is drawn every iteration.
 
 
   Together with the current samples and the chosen samples
 
   Together with the current samples and the chosen samples
 
   -->
 
   -->
   <Option key="debug" value="on"/>
+
   <Option key="debug" value="off"/>
 +
</[[Config:SampleSelector|SampleSelector]]>
 +
</source>
 +
=== extremaLOLA ===
 +
LOLA-Voronoi sample selector supplemented with 1 sample at the minimum and maximum
 +
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 +
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#CombinedSampleSelector|CombinedSampleSelector]]" combineOutputs="false">
 +
  <!-- A highly adaptive sampling algorithm, error and density based -->
 +
  <[[Config:SampleSelector|SampleSelector]] weight="0.8">lola-voronoi</[[Config:SampleSelector|SampleSelector]]>
 +
  <[[Config:SampleSelector|SampleSelector]] weight="0.1">sampleMinimum</[[Config:SampleSelector|SampleSelector]]>
 +
  <[[Config:SampleSelector|SampleSelector]] weight="0.1">sampleMaximum</[[Config:SampleSelector|SampleSelector]]>
 +
 
 +
  <[[Config:MergeCriterion|MergeCriterion]] type="[[MergeCriterion#ClosenessThreshold|ClosenessThreshold]]">
 +
 
 +
      <!-- Closeness threshold, Double -->
 +
      <Option key="closenessThreshold" value="0.05"/>
 +
      <!-- Set a % of the maximumSamples to randomly chosen -->
 +
      <Option key="randomPercentage" value="0"/>
 +
 
 +
      <Option key="debug" value="off"/>
 +
  </[[Config:MergeCriterion|MergeCriterion]]>
 +
</[[Config:SampleSelector|SampleSelector]]>
 +
</source>
 +
=== sampleMinimum ===
 +
Selects one sample at the minimum of the model.
 +
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 +
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#OptimizeCriterion|OptimizeCriterion]]" combineOutputs="false">
 +
  <[[Config:Optimizer|Optimizer]]>patternsearch</[[Config:Optimizer|Optimizer]]>
 +
  <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#minmodel|minmodel]]" scaling="none"/>
 +
</[[Config:SampleSelector|SampleSelector]]>
 +
</source>
 +
=== sampleMaximum ===
 +
Selects one sample at the maximum of the model.
 +
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 +
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#OptimizeCriterion|OptimizeCriterion]]" combineOutputs="false">
 +
  <[[Config:Optimizer|Optimizer]]>patternsearch</[[Config:Optimizer|Optimizer]]>
 +
  <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#maxmodel|maxmodel]]" scaling="none"/>
 +
</[[Config:SampleSelector|SampleSelector]]>
 +
</source>
 +
=== default ===
 +
LOLA sample selector combined with error based sample selector
 +
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 +
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#CombinedSampleSelector|CombinedSampleSelector]]" combineOutputs="false">
 +
  <[[Config:SampleSelector|SampleSelector]] weight="0.7">lola-voronoi</[[Config:SampleSelector|SampleSelector]]>
 +
  <[[Config:SampleSelector|SampleSelector]] weight="0.3">error</[[Config:SampleSelector|SampleSelector]]>
 +
 
 +
  <[[Config:MergeCriterion|MergeCriterion]] type="[[MergeCriterion#ClosenessThreshold|ClosenessThreshold]]"> 
 +
      <!-- Closeness threshold, Double -->
 +
      <Option key="closenessThreshold" value="0.05"/>
 +
      <!-- Set a % of the maximumSamples to randomly chosen -->
 +
      <Option key="randomPercentage" value="0"/>
 +
 
 +
      <Option key="debug" value="off"/>
 +
  </[[Config:MergeCriterion|MergeCriterion]]>
 
</[[Config:SampleSelector|SampleSelector]]>
 
</[[Config:SampleSelector|SampleSelector]]>
 
</source>
 
</source>

Latest revision as of 16:23, 27 February 2014

Generated for SUMO toolbox version 7.0. We are well aware that documentation is not always complete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower. We are are a university research group after all. The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files. If something is unclear please dont hesitate to ask.

SampleSelector

empty

Don't select any new samples, useful when modeling multiple outputs, and you don't want to involve one of these outputs in the sampling process.

<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#EmptySampleSelector|EmptySampleSelector]]" combineOutputs="false"/>

random

Selects new samples randomly in the design space.

<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#RandomSampleSelector|RandomSampleSelector]]" combineOutputs="false"/>

delaunay

This sample selector uses a Delaunay triangulation of the data to select samples in locations far from previous samples, or in locations where the estimated model error is largest. This algorithm uses QHull, which is very slow for high dimensions, so you should only use this sample selector for less than 6D and for less than 1000 samples.

<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#PipelineSampleSelector|PipelineSampleSelector]]" combineOutputs="false">
   
   <[[Config:CandidateGenerator|CandidateGenerator]] type="[[CandidateGenerator#DelaunayCandidateGenerator|DelaunayCandidateGenerator]]"/>
   
        <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#modelDifference|modelDifference]]">
           <Option key="criterion_parameter" value="2"/>
        </[[Config:CandidateRanker|CandidateRanker]]>
        <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#delaunayVolume|delaunayVolume]]"/>
        
        <[[Config:MergeCriterion|MergeCriterion]] type="[[MergeCriterion#WeightedAverage|WeightedAverage]]" weights="[1 1]"/>
        
</[[Config:SampleSelector|SampleSelector]]>

density

A space-filling sampling algorithm which uses an approximation of the Voronoi tessellation of the design space. Will only sample within the "allowed" areas if constraints are specified.

<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#VoronoiSampleRanker|VoronoiSampleRanker]]" combineOutputs="false"/>

error

An adaptive sample selection algorithm (error based), driven by the evaluation of your model on a dense grid, which selects samples in locations where the model error is estimated to be the largest.

<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#PipelineSampleSelector|PipelineSampleSelector]]" combineOutputs="false">

   <[[Config:CandidateGenerator|CandidateGenerator]] type="[[CandidateGenerator#GridCandidateGenerator|GridCandidateGenerator]]"/>
   
   <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#modelDifference|modelDifference]]">
           <Option key="criterion_parameter" value="4"/>
        </[[Config:CandidateRanker|CandidateRanker]]>
        
        <[[Config:MergeCriterion|MergeCriterion]] type="[[MergeCriterion#ClosenessThreshold|ClosenessThreshold]]">
   
      <!-- Closeness threshold, Double -->
      <Option key="closenessThreshold" value="0.05"/>
      <!-- Set a % of the maximumSamples to randomly chosen -->
      <Option key="randomPercentage" value="20"/>
   
      <Option key="debug" value="off"/>
   </[[Config:MergeCriterion|MergeCriterion]]>
</[[Config:SampleSelector|SampleSelector]]>

lola-voronoi

A highly adaptive sampling algorithm which performs a trade-off between exploration (filling up the design space as equally as possible) and exploitation (selecting data points in highly nonlinear regions). lola-voronoi is the only sample selector which currently supports multiple outputs, auto-sampled inputs and constraints.

<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#LOLAVoronoiSampleSelector|LOLAVoronoiSampleSelector]]" combineOutputs="false">
   <!-- Integer between 2 and 20 -->
   <Option key="neighbourhoodSize" value="2"/>
   <!-- Number of frequency values returned for each submitted sample. Only used with auto-sampled inputs. -->
   <Option key="frequencies" value="11"/>
</[[Config:SampleSelector|SampleSelector]]>

rationalPoleSupression

A sampling algorithm aimed at supressing poles in rational models by sampling them (only for Rational models)

<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#OptimizeCriterion|OptimizeCriterion]]" combineOutputs="false">

   <!-- This criterion has to be solved to choose new samples, one can choose the optimizer used here -->
   <[[Config:Optimizer|Optimizer]]>patternsearch</[[Config:Optimizer|Optimizer]]>
   
   <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#rationalPoleSupression|rationalPoleSupression]]" scaling="none"/>
   <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#modelDifference|modelDifference]]" scaling="none"/>

   <!--
   when debug is 'on' a contour plot of the criterion function is drawn every iteration.
   Together with the current samples and the chosen samples
   -->
   <Option key="debug" value="off"/>
</[[Config:SampleSelector|SampleSelector]]>

expectedImprovement

A sampling algorithm aimed at optimization problems (only for Kriging and RBF)

<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#OptimizeCriterion|OptimizeCriterion]]" combineOutputs="false">

   <!-- This criterion has to be solved to choose new samples, one can choose the optimizer used here -->
   <[[Config:Optimizer|Optimizer]]>patternsearch</[[Config:Optimizer|Optimizer]]>
   
   <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#expectedImprovement|expectedImprovement]]" scaling="none">
   </[[Config:CandidateRanker|CandidateRanker]]>
   <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#maxvar|maxvar]]" scaling="none"/>

   <!--
   when debug is 'on' a contour plot of the criterion function is drawn every iteration.
   Together with the current samples and the chosen samples
   -->
   <Option key="debug" value="off"/>
</[[Config:SampleSelector|SampleSelector]]>

extremaLOLA

LOLA-Voronoi sample selector supplemented with 1 sample at the minimum and maximum

<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#CombinedSampleSelector|CombinedSampleSelector]]" combineOutputs="false">
   <!-- A highly adaptive sampling algorithm, error and density based -->
   <[[Config:SampleSelector|SampleSelector]] weight="0.8">lola-voronoi</[[Config:SampleSelector|SampleSelector]]>
   <[[Config:SampleSelector|SampleSelector]] weight="0.1">sampleMinimum</[[Config:SampleSelector|SampleSelector]]>
   <[[Config:SampleSelector|SampleSelector]] weight="0.1">sampleMaximum</[[Config:SampleSelector|SampleSelector]]>
   
   <[[Config:MergeCriterion|MergeCriterion]] type="[[MergeCriterion#ClosenessThreshold|ClosenessThreshold]]">
   
      <!-- Closeness threshold, Double -->
      <Option key="closenessThreshold" value="0.05"/>
      <!-- Set a % of the maximumSamples to randomly chosen -->
      <Option key="randomPercentage" value="0"/>
   
      <Option key="debug" value="off"/>
   </[[Config:MergeCriterion|MergeCriterion]]>
</[[Config:SampleSelector|SampleSelector]]>

sampleMinimum

Selects one sample at the minimum of the model.

<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#OptimizeCriterion|OptimizeCriterion]]" combineOutputs="false">
   <[[Config:Optimizer|Optimizer]]>patternsearch</[[Config:Optimizer|Optimizer]]>
   <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#minmodel|minmodel]]" scaling="none"/>
</[[Config:SampleSelector|SampleSelector]]>

sampleMaximum

Selects one sample at the maximum of the model.

<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#OptimizeCriterion|OptimizeCriterion]]" combineOutputs="false">
   <[[Config:Optimizer|Optimizer]]>patternsearch</[[Config:Optimizer|Optimizer]]>
   <[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#maxmodel|maxmodel]]" scaling="none"/>
</[[Config:SampleSelector|SampleSelector]]>

default

LOLA sample selector combined with error based sample selector

<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#CombinedSampleSelector|CombinedSampleSelector]]" combineOutputs="false">
   <[[Config:SampleSelector|SampleSelector]] weight="0.7">lola-voronoi</[[Config:SampleSelector|SampleSelector]]>
   <[[Config:SampleSelector|SampleSelector]] weight="0.3">error</[[Config:SampleSelector|SampleSelector]]>
   
   <[[Config:MergeCriterion|MergeCriterion]] type="[[MergeCriterion#ClosenessThreshold|ClosenessThreshold]]">   
      <!-- Closeness threshold, Double -->
      <Option key="closenessThreshold" value="0.05"/>
      <!-- Set a % of the maximumSamples to randomly chosen -->
      <Option key="randomPercentage" value="0"/>
   
      <Option key="debug" value="off"/>
   </[[Config:MergeCriterion|MergeCriterion]]>
</[[Config:SampleSelector|SampleSelector]]>