Difference between revisions of "Config:DataSource"

From SUMOwiki
Jump to navigationJump to search
Line 43: Line 43:
  
 
=== griddedDataset ===
 
=== griddedDataset ===
Evaluate samples using a gridded dataset
+
Evaluate samples using a gridded dataset. This data format does not include any inputs, but lists only outputs and only work for a uniform grid of data points. The order in which the output values are given determine their location in the grid.
 +
 
 +
For example, if you want to define a 3-dimensional dataset with grid size 2x3x2 on the [-1,1] domain, you must provide the outputs for the samples in the following order:
 +
 
 +
<code><pre>
 +
value at [-1, -1, -1]
 +
value at [-1, -1,  1]
 +
value at [-1,  0, -1]
 +
value at [-1,  0,  1]
 +
value at [-1,  1, -1]
 +
value at [-1,  1,  1]
 +
value at [ 1, -1, -1]
 +
value at [ 1, -1,  1]
 +
value at [ 1,  0, -1]
 +
value at [ 1,  0,  1]
 +
value at [ 1,  1, -1]
 +
value at [ 1,  1,  1]
 +
</pre></code>
 +
 
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 
<source xmlns:saxon="http://icl.com/saxon" lang="xml">
 
<[[Config:SampleEvaluator|SampleEvaluator]] id="griddedDataset" type="ibbt.sumo.sampleevaluators.datasets.GriddedDatasetSampleEvaluator">
 
<[[Config:SampleEvaluator|SampleEvaluator]] id="griddedDataset" type="ibbt.sumo.sampleevaluators.datasets.GriddedDatasetSampleEvaluator">
Line 50: Line 68:
 
</[[Config:SampleEvaluator|SampleEvaluator]]>
 
</[[Config:SampleEvaluator|SampleEvaluator]]>
 
</source>
 
</source>
 +
 
=== scatteredDataset ===
 
=== scatteredDataset ===
 
Evaluate samples using a scattered dataset. Each row of the dataset represents a data point. If the dimensionality of the problem is D (i.e. there are D inputs), the first D columns represent the inputs and the remaining columns the outputs. Note that complex number need to given as separate columns.
 
Evaluate samples using a scattered dataset. Each row of the dataset represents a data point. If the dimensionality of the problem is D (i.e. there are D inputs), the first D columns represent the inputs and the remaining columns the outputs. Note that complex number need to given as separate columns.

Revision as of 17:50, 22 February 2012

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 We are well aware that documentation is not always complete and possibly even out of date in some cases.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.

SampleEvaluator

local

Use this if you data generator is a native executable, shell script, or java class

<[[Config:SampleEvaluator|SampleEvaluator]] id="local" type="ibbt.sumo.sampleevaluators.LocalSampleEvaluator">
   <!-- Maximum number of times to resubmit a point (e.g., in case something went wrong) -->
   <Option key="maxResubmissions" value="1"/>
   <!-- If a sample takes longer than "sampleTimeout*average evaluation time of one sample" 
   seconds to evaluate it is removed from the pending list (set to -1 to disable) -->
   <Option key="sampleTimeout" value="-1"/>
   <!-- Can be set to "java" for java executables, to "external" for platform-specific
        binaries/scripts and to nothing at all for auto-detection.-->
   <Option key="simulatorType" value=""/>
   <!-- Can be set to a number of seconds, if one simulator evaluation exceeds this timeout,
        the simulation is aborted -->
   <!-- <Option key="timeout" value="12"/> -->
   <!-- Number of samples to evaluate concurrently, useful if you have a dual or multi-core machine -->
   <Option key="threadCount" value="1"/>
</[[Config:SampleEvaluator|SampleEvaluator]]>

matlab

Evaluate samples using a matlab script (ie. your simulator is a matlab script)

<[[Config:SampleEvaluator|SampleEvaluator]] id="matlab" type="ibbt.sumo.sampleevaluators.matlab.MatlabSampleEvaluator">
   <!-- Maximum number of times to resubmit a point (e.g., in case something went wrong) -->
   <Option key="maxResubmissions" value="1"/>
   <!-- If a sample takes longer than "sampleTimeout*average evaluation time of one sample" 
   seconds to evaluate it is removed from the pending list (set to -1 to disable) -->
   <Option key="sampleTimeout" value="-1"/>
</[[Config:SampleEvaluator|SampleEvaluator]]>


matlabDirect

Evaluate samples using a matlab script without using Java.

	<SampleEvaluator id="matlabDirect" type="MatlabDirectSampleEvaluator"/>

griddedDataset

Evaluate samples using a gridded dataset. This data format does not include any inputs, but lists only outputs and only work for a uniform grid of data points. The order in which the output values are given determine their location in the grid.

For example, if you want to define a 3-dimensional dataset with grid size 2x3x2 on the [-1,1] domain, you must provide the outputs for the samples in the following order:

value at [-1, -1, -1]
value at [-1, -1,  1]
value at [-1,  0, -1]
value at [-1,  0,  1]
value at [-1,  1, -1]
value at [-1,  1,  1]
value at [ 1, -1, -1]
value at [ 1, -1,  1]
value at [ 1,  0, -1]
value at [ 1,  0,  1]
value at [ 1,  1, -1]
value at [ 1,  1,  1]
<[[Config:SampleEvaluator|SampleEvaluator]] id="griddedDataset" type="ibbt.sumo.sampleevaluators.datasets.GriddedDatasetSampleEvaluator">
   <!-- Using an ID you can specify which dataset from the simulator file to use -->
   <!-- <Option key="id" value="someDataset"/> -->
</[[Config:SampleEvaluator|SampleEvaluator]]>

scatteredDataset

Evaluate samples using a scattered dataset. Each row of the dataset represents a data point. If the dimensionality of the problem is D (i.e. there are D inputs), the first D columns represent the inputs and the remaining columns the outputs. Note that complex number need to given as separate columns.

<[[Config:SampleEvaluator|SampleEvaluator]] id="scatteredDataset" type="ibbt.sumo.sampleevaluators.datasets.ScatteredDatasetSampleEvaluator">
   <!-- Using an ID you can specify which dataset from the simulator file to use -->
   <!-- <Option key="id" value="someDataset"/> -->
</[[Config:SampleEvaluator|SampleEvaluator]]>

calcua

Evaluate samples on a SGE administered cluster through a remote, ssh reachable frontnode

<[[Config:SampleEvaluator|SampleEvaluator]] id="calcua" type="ibbt.sumo.sampleevaluators.distributed.sge.RemoteSGESampleEvaluator">
   <!-- Maximum number of times to resubmit a point (e.g., in case something went wrong) -->
   <Option key="maxResubmissions" value="1"/>
   <!-- If a sample takes longer than "sampleTimeout*average evaluation time of one sample" 
        seconds to evaluate it is removed from the pending list (set to -1 to disable) -->
   <Option key="sampleTimeout" value="-1"/>
   
   <!--The platform specs of the cluster-->
   <[[Config:Executable|Executable]] platform="linux" arch="x86_64"/>
   <[[Config:Backend|Backend]] id="remoteSGE" type="ibbt.sumo.sampleevaluators.distributed.sge.RemoteSGEBackend">
      <!--ssh user name to login on the front node-->
      <Option key="user" value="dgorisse"/>
      <!--Submissions happen from this front node, you need to have key-based ssh authentication-->
      <Option key="frontNode" value="submit.calcua.ua.ac.be"/>
      <!--Directory on the front node where input/output files, dependencies, etc. are stored-->
      <Option key="remoteDirectory" value="/storeA/users/dgorisse/output"/>
      <!--poll for result files every xx seconds-->
      <Option key="pollInterval" value="20"/>
      <!--queues we can submit to-->
      <Option key="queues" value="all.q,fast.q"/>
      <!--check for faster queues (more slots available) every xx seconds-->
      <Option key="queueRevisionRate" value="10"/>
      <!--script on the frontnode that sets up the necessary shell environment-->
      <Option key="environmentCommand" value=". ~/.profile;"/>
   </[[Config:Backend|Backend]]>
</[[Config:SampleEvaluator|SampleEvaluator]]>