#include <AbstractBestSubsets.hh>

Inheritance diagram for AbstractBestSubsets:

Collaboration diagram for AbstractBestSubsets:

Public Member Functions
	AbstractBestSubsets (AlgMetadata const *)
virtual	~AbstractBestSubsets ()=0
int	needNormalization ()
int	initialize ()
int	iterate ()
int	done () const
float	getProgress () const
Scalar	getValue (const Sample &x) const
int	getConvergence (Scalar *const val) const
Protected Member Functions
virtual void	_getConfiguration (ConfigurationPtr &) const
virtual void	_setConfiguration (const ConstConfigurationPtr &)
Protected Attributes
std::string	_subAlgorithm
Private Member Functions
int	numActiveThreads ()
int	earlyTerminationConditionMet ()
int	calculateBestSubset ()
void	sortRuns (AlgorithmRun **runList, int nelements, int errorType)
Private Attributes
double	_trainProp
int	_totalRuns
double	_omissionThreshold
int	_modelsUnderOmission
double	_commissionThreshold
int	_commissionSampleSize
int	_maxThreads
bool	_softOmissionThreshold
int	_currentModelsUnderOmissionThreshold
AlgorithmRun **	_finishedRun
AlgorithmRun **	_activeRun
AlgorithmRun **	_bestRun
int	_numFinishedRuns
int	_numActiveRuns
int	_numBestRuns
int	_done
float	_maxProgress

Detailed Description

Declaration of Best Subsets Procedure base class

Author:: Ricardo Scachetti Pereira (rpereira@ku.edu)

Date:: 2004-12-10

Id:: AbstractBestSubsets.hh 5932 2014-03-17 22:51:58Z rdg

LICENSE INFORMATION

Copyright(c), The Center for Research, University of Kansas, 2385 Irving Hill Road, Lawrence, KS 66044-4755, USA.

http://www.nhm.ku.edu

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details:

http://www.gnu.org/copyleft/gpl.html Implements Best Subset Procedure (Anderson et al. 2003).

It runs "n" Garp models, filter the best ones based on the procedure defined by Anderson et al. 2003 and sum them to obtain a probability map for the species distribution.

Current implementation uses multi-threading to run multiple Garp runs in parallel. Takes advantage of multi-processor servers and workstations.

Definition at line 53 of file AbstractBestSubsets.hh.

Constructor & Destructor Documentation

AbstractBestSubsets::AbstractBestSubsets ( AlgMetadata const * metadata_bs )

Definition at line 64 of file AbstractBestSubsets.cpp.

References _activeRun, _bestRun, _commissionSampleSize, _commissionThreshold, _currentModelsUnderOmissionThreshold, _done, _finishedRun, _maxProgress, _maxThreads, _modelsUnderOmission, _numActiveRuns, _numFinishedRuns, _omissionThreshold, _softOmissionThreshold, _totalRuns, and _trainProp.

AbstractBestSubsets::~AbstractBestSubsets ( ) [pure virtual]

Definition at line 91 of file AbstractBestSubsets.cpp.

References _activeRun, _bestRun, _finishedRun, and _numFinishedRuns.

Member Function Documentation

void AbstractBestSubsets::_getConfiguration ( ConfigurationPtr & config ) const [protected, virtual]

Reimplemented from AlgorithmImpl.

Definition at line 114 of file AbstractBestSubsets.cpp.

References _bestRun, _done, _numBestRuns, and AlgorithmRun::getAlgorithm().

Here is the call graph for this function:

void AbstractBestSubsets::_setConfiguration ( const ConstConfigurationPtr & config ) [protected, virtual]

Reimplemented from AlgorithmImpl.

Definition at line 141 of file AbstractBestSubsets.cpp.

References _bestRun, _done, _numBestRuns, and AlgorithmFactory::newAlgorithm().

Here is the call graph for this function:

int AbstractBestSubsets::calculateBestSubset ( ) [private]

Definition at line 423 of file AbstractBestSubsets.cpp.

References _bestRun, _commissionThreshold, _finishedRun, _modelsUnderOmission, _numBestRuns, _numFinishedRuns, _totalRuns, Log::info(), Log::instance(), printListOfRuns(), and sortRuns().

Referenced by iterate().

Here is the call graph for this function:

int AbstractBestSubsets::done ( ) const [virtual]

Check if the model generation is completed (e.g. convergence point has been met.

Returns:: Implementation specific but usually 1 for completion.

Reimplemented from AlgorithmImpl.

Definition at line 500 of file AbstractBestSubsets.cpp.

References _done.

Referenced by getProgress().

int AbstractBestSubsets::earlyTerminationConditionMet ( ) [private]

Definition at line 416 of file AbstractBestSubsets.cpp.

References _currentModelsUnderOmissionThreshold, _modelsUnderOmission, and _softOmissionThreshold.

Referenced by iterate().

int AbstractBestSubsets::getConvergence ( Scalar *const val ) const [virtual]

Returns a value that represents the convergence of the algorithm expressed as a number between 0 and 1 where 0 represents model completion.

Returns:

Parameters:

val

Reimplemented from AlgorithmImpl.

Definition at line 560 of file AbstractBestSubsets.cpp.

float AbstractBestSubsets::getProgress ( ) const [virtual]

Return progress so far

Reimplemented from AlgorithmImpl.

Definition at line 508 of file AbstractBestSubsets.cpp.

References _activeRun, _currentModelsUnderOmissionThreshold, _maxProgress, _modelsUnderOmission, _numActiveRuns, _numFinishedRuns, _softOmissionThreshold, _totalRuns, done(), and AlgorithmRun::getProgress().

Here is the call graph for this function:

Scalar AbstractBestSubsets::getValue ( const Sample & x ) const [virtual]

This method is used when projecting the model.

Note:: This method is inherited from the Algorithm class

Returns:

Parameters:

x	Pointer to a vector of openModeller Scalar type (currently double). The vector should contain values looked up on the environmental variable layers into which the mode is being projected.

Implements AlgorithmImpl.

Definition at line 543 of file AbstractBestSubsets.cpp.

References _bestRun, _done, _numBestRuns, and AlgorithmRun::getValue().

Here is the call graph for this function:

int AbstractBestSubsets::initialize ( ) [virtual]

Initialize data structures

Note:: This method is inherited from the Algorithm class

Returns:: 0 on error

Implements AlgorithmImpl.

Definition at line 228 of file AbstractBestSubsets.cpp.

References _activeRun, _commissionSampleSize, _commissionThreshold, _finishedRun, _maxThreads, _modelsUnderOmission, _omissionThreshold, _softOmissionThreshold, _subAlgorithm, _totalRuns, _trainProp, Log::error(), AlgorithmImpl::getParameter(), Log::instance(), and Log::warn().

Here is the call graph for this function:

int AbstractBestSubsets::iterate ( ) [virtual]

Start model execution (build the model). Consists of a single GARP run. It is also the control loop for the threads

Returns:: 0 on error

Reimplemented from AlgorithmImpl.

Definition at line 311 of file AbstractBestSubsets.cpp.

References _activeRun, _commissionSampleSize, _done, _maxThreads, _numActiveRuns, _numFinishedRuns, AlgorithmImpl::_param, AlgorithmImpl::_samp, _subAlgorithm, _totalRuns, _trainProp, calculateBestSubset(), earlyTerminationConditionMet(), AlgorithmRun::initialize(), AlgorithmFactory::newAlgorithm(), numActiveThreads(), AlgorithmRun::run(), SLEEP, and splitSampler().

Here is the call graph for this function:

int AbstractBestSubsets::needNormalization ( ) [virtual]

This method is used when you want to ensure that all variables in all environmental layers are scaled to the same value range. GARP requires values to be normalized between -1.0 and 1.0.