Using a model - Revision history

Mnguyen: /* Model object interfacing and optimization */

2012-02-10T16:28:03Z

Model object interfacing and optimization

← Older revision		Revision as of 16:28, 10 February 2012
Line 155:		Line 155:
	feval( handle, 0, 1, -1 );		feval( handle, 0, 1, -1 );
	</pre></code>		</pre></code>

			== Miscellaneous post-processing scripts ==

			The SUMO Toolbox also includes some miscellaneous scripts you can use to analyze the results of your run. You can find them in <code>...\SUMO-Toolbox\src\scripts\matlab\</code> folder.

Dgorissen: /* Available methods */

2010-08-18T07:23:20Z

Available methods

← Older revision		Revision as of 07:23, 18 August 2010
Line 54:		Line 54:
	>>str = struct(model)		>>str = struct(model)
	</source>		</source>

			Also, some model types provide methods to access the internal model representation. For example, if m is an object of type ANNModel, then executing m.getNetwork() will return the nested Matlab neural network object (from the Matlab Neural Network Toolbox).

	=== guiPlotModel ===		=== guiPlotModel ===

Dgorissen: /* Available methods */

2010-08-18T07:21:05Z

Available methods

← Older revision		Revision as of 07:21, 18 August 2010
Line 48:		Line 48:
	If you want to understand the structure of the model, i.e., how the model object is built up you can do two things:		If you want to understand the structure of the model, i.e., how the model object is built up you can do two things:

	# open the class file for that model (e.g., for an ANNModel object, open the src/matlab/models/@ANNModel/ANNModel.m file		# open the class file for that model, e.g., for an ANNModel object, open the src/matlab/models/@ANNModel/ANNModel.m file
	# use the struct command to convert the model object to a structure. For example:		# use the struct command to convert the model object to a structure. For example:

Line 54:		Line 54:
	>>str = struct(model)		>>str = struct(model)
	</source>		</source>


	=== guiPlotModel ===		=== guiPlotModel ===

Dgorissen: /* Available methods */

2010-08-18T07:20:44Z

Available methods

← Older revision		Revision as of 07:20, 18 August 2010
Line 61:		Line 61:

	=== plotModel ===		=== plotModel ===
	<~~code><pre~~>		<source lang="matlab">
	>>[figureHandle] = plotModel(model,[outputNumber],[options])		>>[figureHandle] = plotModel(model,[outputNumber],[options])
	</~~pre></code~~>		</source>

	<code>plotModel</code> will generate an indicative plot of the model surface. To do so, it evaluates the model on a reasonably dense grid of points.		<code>plotModel</code> will generate an indicative plot of the model surface. To do so, it evaluates the model on a reasonably dense grid of points.
Line 83:		Line 83:

	=== evaluate ===		=== evaluate ===
	<~~code><pre~~>		<source lang="matlab">
	>> values = evaluate(model, samples);		>> values = evaluate(model, samples);
	</~~pre></code~~>		</source>

	This evaluates the model on the given samples. The samples should be provided in simulator space. Simulator space is defined by the range in the [[Simulator configuration]]. If no range (minimum and maximum) was specified, the domain is assumed to be [-1,1].		This evaluates the model on the given samples. The samples should be provided in simulator space. Simulator space is defined by the range in the [[Simulator configuration]]. If no range (minimum and maximum) was specified, the domain is assumed to be [-1,1].
Line 92:		Line 92:

	=== evaluateDerivative ===		=== evaluateDerivative ===
	<~~code><pre~~>		<source lang="matlab">
	>> values = evaluateDerivative(model, samples, [outputIndex]);		>> values = evaluateDerivative(model, samples, [outputIndex]);
	</~~pre></code~~>		</source>

	This approximates the partial derivatives of the model at each given sample. Note that the base class implementation is a very simple approximation. Models can override this function to provide more accurate derivatives (e.g., Kriging does this already). However, in its current form it is already useful.		This approximates the partial derivatives of the model at each given sample. Note that the base class implementation is a very simple approximation. Models can override this function to provide more accurate derivatives (e.g., Kriging does this already). However, in its current form it is already useful.

	=== getSamples ===		=== getSamples ===
	<~~code><pre~~>		<source lang="matlab">
	>> samples = getSamples(model);		>> samples = getSamples(model);
	</~~pre></code~~>		</source>

	Returns the samples that were used to fit the model. The samples are returned in simulator space.		Returns the samples that were used to fit the model. The samples are returned in simulator space.

	=== getValues ===		=== getValues ===
	<~~code><pre~~>		<source lang="matlab">
	>> values = getValues(model);		>> values = getValues(model);
	</~~pre></code~~>		</source>

	Returns the values that correspond to the samples from getSamples().		Returns the values that correspond to the samples from getSamples().

	=== getDescription ===		=== getDescription ===
	<~~code><pre~~>		<source lang="matlab">
	>> desc = getDescription(model);		>> desc = getDescription(model);
	</~~pre></code~~>		</source>

	Returns a string with a user friendly description of the model.		Returns a string with a user friendly description of the model.

	=== getExpression ===		=== getExpression ===
	<~~code><pre~~>		<source lang="matlab">
	>> desc = getExpression(model,[outputNumber]);		>> desc = getExpression(model,[outputNumber]);
	</~~pre></code~~>		</source>

	Returns the symbolic mathematical expression of this model (e.g., 3x1^2 - 2x2 +5). Note that not all model types implement this.		Returns the symbolic mathematical expression of this model (e.g., 3x1^2 - 2x2 +5). Note that not all model types implement this.

	=== construct ===		=== construct ===
	<~~code><pre~~>		<source lang="matlab">
	>> model = construct(model,samples);		>> model = construct(model,samples);
	</~~pre></code~~>		</source>

	This will build/train/fit.. the model on the given set of data points and return the updated model.		This will build/train/fit.. the model on the given set of data points and return the updated model.

	=== complexity ===		=== complexity ===
	<~~code><pre~~>		<source lang="matlab">
	>> n = complexity(model);		>> n = complexity(model);
	</~~pre></code~~>		</source>

	Returns the number of free parameters in the model. By default this returns the number of datapoints the model was built with but this is overridden by some model types. For example, an ANN model returns the number of weights in the network while a rational model returns the number of coefficients.		Returns the number of free parameters in the model. By default this returns the number of datapoints the model was built with but this is overridden by some model types. For example, an ANN model returns the number of weights in the network while a rational model returns the number of coefficients.

Dgorissen: /* Available methods */

2010-08-18T07:20:17Z

Available methods

← Older revision		Revision as of 07:20, 18 August 2010
Line 46:		Line 46:
	</source>		</source>

			If you want to understand the structure of the model, i.e., how the model object is built up you can do two things:

			# open the class file for that model (e.g., for an ANNModel object, open the src/matlab/models/@ANNModel/ANNModel.m file
			# use the struct command to convert the model object to a structure. For example:

			<source lang="matlab">
			>>str = struct(model)
			</source>

Dgorissen: /* Available methods */

2010-08-18T07:16:31Z

Available methods

← Older revision		Revision as of 07:16, 18 August 2010
Line 44:		Line 44:
	<source lang="matlab">		<source lang="matlab">
	>>methods(model)		>>methods(model)
	</~~matlab~~>		</source>

Dgorissen: /* Available methods */

2010-08-18T07:16:24Z

Available methods

← Older revision		Revision as of 07:16, 18 August 2010
Line 39:		Line 39:

	== Available methods ==		== Available methods ==

	~~'''Note this discussion refers to the SUMO-Toolbox version 6.1'''~~

	Once the model is loaded you can invoke a number methods on it. We list the main ones below. For a full list of available methods just use the matlab 'methods' command:		Once the model is loaded you can invoke a number methods on it. We list the main ones below. For a full list of available methods just use the matlab 'methods' command:

	<~~code><pre~~>		<source lang="matlab">
	>>methods(model)		>>methods(model)
	</~~pre></code~~>		</matlab>



	=== guiPlotModel ===		=== guiPlotModel ===

Dgorissen: /* freeParams */

2009-09-29T19:27:41Z

freeParams

← Older revision		Revision as of 19:27, 29 September 2009
Line 125:		Line 125:
	This will build/train/fit.. the model on the given set of data points and return the updated model.		This will build/train/fit.. the model on the given set of data points and return the updated model.

	=== ~~freeParams~~ ===		=== complexity ===
	<code><pre>		<code><pre>
	>> n = ~~freeParams~~(model);		>> n = complexity(model);
	</pre></code>		</pre></code>

Digipham: orthography fixed

2009-03-04T09:39:53Z

orthography fixed

← Older revision		Revision as of 09:39, 4 March 2009
Line 134:		Line 134:
	== Model object interfacing and optimization ==		== Model object interfacing and optimization ==

	You may want to use the model as part of a larger Matlab program, or you may simply want to optimizer the model. To easily do this you can create a function handle to the model object. You can ~~doe~~ this as follows (example for the 3D case):		You may want to use the model as part of a larger Matlab program, or you may simply want to optimizer the model. To easily do this you can create a function handle to the model object. You can do this as follows (example for the 3D case):

	<code><pre>		<code><pre>

Dgorissen: /* Model space vs Simulator space */

2009-03-02T14:51:45Z

Model space vs Simulator space

← Older revision		Revision as of 14:51, 2 March 2009
Line 36:		Line 36:
	It is important to note the difference between '''Model space''' and '''Simulator space'''. When a data point is in model space, it means its inputs all lie in the range [-1 1]. When the point is in simulator space its inputs lie in the range specified by the [[Simulator configuration]] file.		It is important to note the difference between '''Model space''' and '''Simulator space'''. When a data point is in model space, it means its inputs all lie in the range [-1 1]. When the point is in simulator space its inputs lie in the range specified by the [[Simulator configuration]] file.

	Internally the toolbox only works in model space. The toolbox will take care of translating points from simulator space into model space and back (this happens in the SampleManager object). You will note that many methods have a ''XXXinModelSpace'' variant. This just means that the method does exactly the same, except it expects points to be in model space. You should normally not have to care about model space unless you are writing your own extensions to the toolbox. In that case see [[Add ~~model type~~]].		Internally the toolbox only works in model space. The toolbox will take care of translating points from simulator space into model space and back (this happens in the SampleManager object). You will note that many methods have a ''XXXinModelSpace'' variant. This just means that the method does exactly the same, except it expects points to be in model space. You should normally not have to care about model space unless you are writing your own extensions to the toolbox. In that case see [[Add Model Type]].

	== Available methods ==		== Available methods ==