#include <OpenModeller.hh>

Collaboration diagram for OpenModeller:

Public Member Functions
	OpenModeller ()

	~OpenModeller ()

void	setLogLevel (Log::Level level)

std::string	getVersion ()

AlgMetadata const **	availableAlgorithms ()

AlgMetadata const *	algorithmMetadata (char const *algorithm_id)

int	numAvailableAlgorithms ()

EnvironmentPtr	getEnvironment ()

AlgorithmPtr	getAlgorithm ()

Model	getModel () const

const SamplerPtr &	getSampler () const

int	setOccurrences (const OccurrencesPtr &presence, const OccurrencesPtr &absence=OccurrencesPtr())

int	setAlgorithm (std::string const id, int nparam, AlgParameter const *param)

void	setEnvironment (std::vector< std::string > categ_map, std::vector< std::string > continuous_map, const std::string &mask)

void	setSampler (const SamplerPtr &sampler)

void	setAbortionCallback (AbortionCallback func, void *param=0)

void	setMapCallback (ModelProjectionCallback func, void *param=0)

int	createMap (const EnvironmentPtr &env, char const *output_file, MapFormat &format)

int	createMap (const EnvironmentPtr &env, char const *output_file)

int	createMap (char const *output_file, MapFormat &format)

int	createMap (char const *output_file)

void	setModelCallback (ModelCreationCallback func, void *param=0)

int	createModel ()

int	run ()

Scalar	getValue (const ConstEnvironmentPtr &env, Coord x, Coord y)

Scalar	getValue (Scalar const *environment_values)

char *	error ()

AreaStats *	getActualAreaStats ()

AreaStats *	getEstimatedAreaStats (double proportionAreaToSample=0.01)

AreaStats *	getEstimatedAreaStats (const ConstEnvironmentPtr &env, double proportionAreaToSample=0.01)

const ConfusionMatrix *const	getConfusionMatrix ()

RocCurve *const	getRocCurve ()

ConfigurationPtr	getModelConfiguration () const

void	setModelConfiguration (const ConstConfigurationPtr &)

void	setProjectionConfiguration (const ConstConfigurationPtr &)

void	calculateModelStatistics (const ConstConfigurationPtr &)

Private Member Functions
bool	hasEnvironment ()

Private Attributes
SamplerPtr	_samp

AlgorithmPtr	_alg

OccurrencesPtr	_presence

OccurrencesPtr	_absence

EnvironmentPtr	_env

CallbackWrapper	_callback_wrapper

MapFormat	_format

EnvironmentPtr	_projEnv

AreaStats *	_actualAreaStats

AreaStats *	_estimatedAreaStats

ConfusionMatrix	_confusion_matrix

RocCurve	_roc_curve

char	_error [256]

Detailed Description

Defines and implements all commands to interface with the model generator.

Definition at line 71 of file OpenModeller.hh.

Constructor & Destructor Documentation

OpenModeller::OpenModeller ( )

Definition at line 76 of file OpenModeller.cpp.

References _actualAreaStats, _error, and _estimatedAreaStats.

OpenModeller::~OpenModeller ( )

Definition at line 90 of file OpenModeller.cpp.

References _actualAreaStats, and _estimatedAreaStats.

Member Function Documentation

AlgMetadata const * OpenModeller::algorithmMetadata ( char const * algorithm_id )

Returns a specific algorithm metadata

Parameters

algorithm_id Identifier of the algorithm.

Returns: Algorithm's metadata or zero if the algorithm was not found.

Definition at line 126 of file OpenModeller.cpp.

References AlgorithmFactory::algorithmMetadata().

Referenced by RequestFile::_setAlgorithm().

Here is the call graph for this function:

AlgMetadata const ** OpenModeller::availableAlgorithms ( )

Finds the system available algorithms' metadata.

The pointer returned are copied from an internal storage of algorithms. So they can not be deallocated. Another point is that the AlgMetadata will be reallocated the next time this method is called.

Returns: a null terminated list of available algorithms.

Definition at line 117 of file OpenModeller.cpp.

References AlgorithmFactory::availableAlgorithms().

Referenced by main().

Here is the call graph for this function:

void OpenModeller::calculateModelStatistics ( const ConstConfigurationPtr & config )

Definition at line 648 of file OpenModeller.cpp.

References _alg, _confusion_matrix, _roc_curve, _samp, RocCurve::calculate(), ConfusionMatrix::calculate(), CONF_MATRIX_DEFAULT_THRESHOLD, Log::debug(), Log::error(), getModel(), RocCurve::getPartialAreaRatio(), getSampler(), ConfusionMatrix::getThreshold(), RocCurve::getTotalArea(), Log::info(), RocCurve::initialize(), Log::instance(), ConfusionMatrix::reset(), RocCurve::reset(), ROC_DEFAULT_RESOLUTION, ConfusionMatrix::setLowestTrainingThreshold(), and UNUSED.

Referenced by main().

Here is the call graph for this function:

int OpenModeller::createMap	(	const EnvironmentPtr &	env,
		char const *	output_file,
		MapFormat &	format
	)

Create and save distribution map to disk using the specified projection environment and output format.

Parameters

env	Pointer to Environment object with the layers to project the model onto.
output_file	Output file name.
format	Georeferenced map file which will define cell size, extent, WKT projection, no data value, and file type for the output map.

Definition at line 265 of file OpenModeller.cpp.

References _actualAreaStats, _alg, _callback_wrapper, _format, _projEnv, MapFormat::copyDefaults(), Projector::createMap(), Log::error(), MapFormat::GreyBMP, Log::info(), Log::instance(), RasterFactory::instance(), MapFormat::setFormat(), and Log::warn().

Referenced by createMap(), main(), and RequestFile::makeProjection().

Here is the call graph for this function:

int OpenModeller::createMap	(	const EnvironmentPtr &	env,
		char const *	output_file
	)

Create and save distribution map to disk using the specified projection environment. Output format defaults to the output mask format.

Parameters

env	Pointer to Environment object with the layers to project the model onto.
output_file	Output file name.

Definition at line 335 of file OpenModeller.cpp.

References _format, and createMap().

Here is the call graph for this function:

int OpenModeller::createMap	(	char const *	output_file,
		MapFormat &	format
	)

Create and save distribution map to disk using the specified output format. Projection environment defaults to the same environment used during model creation and previously set by calling the setEnvironment() method.

Parameters

output_file	Output file name.
format	Georeferenced map file which will define cell size, extent, WKT projection, no data value, and file type for the output map.

Definition at line 356 of file OpenModeller.cpp.

References _env, _projEnv, and createMap().

Here is the call graph for this function:

int OpenModeller::createMap ( char const * output_file )

Create and save distribution map to disk. Projection environment defaults to the same environment used during model creation and previously set by calling the setEnvironment() method. Output format defaults to the output mask format.

Parameters

output_file Output file name.

Definition at line 343 of file OpenModeller.cpp.

References _env, _format, _projEnv, and createMap().

Here is the call graph for this function:

int OpenModeller::createModel ( )

Run the algorithm to create the model.

Definition at line 231 of file OpenModeller.cpp.

References _alg, _callback_wrapper, _confusion_matrix, _roc_curve, _samp, Log::error(), Log::info(), Log::instance(), ConfusionMatrix::reset(), and RocCurve::reset().

Referenced by main(), and RequestFile::makeModel().

Here is the call graph for this function:

char* OpenModeller::error ( )

inline

Definition at line 304 of file OpenModeller.hh.

Referenced by RequestFile::makeModel().

AreaStats * OpenModeller::getActualAreaStats ( )

Returns a pointer to the model AreaStats object which contains statistics about areas on the map generated by OM. IMPORTANT: you should delete the pointer because it returns a copy of the internal object!

Definition at line 405 of file OpenModeller.cpp.

References _actualAreaStats.

Referenced by main().

AlgorithmPtr OpenModeller::getAlgorithm ( )

inline

Returns the current algorithm setting.

Returns: Pointer to algorithm.

Definition at line 139 of file OpenModeller.hh.

Referenced by main().

const ConfusionMatrix *const OpenModeller::getConfusionMatrix ( )

Returns a pointer to the internal object representing the confusion matrix that was calculated from the last generated model. IMPORTANT: you should NOT delete the returned pointer!

Definition at line 462 of file OpenModeller.cpp.

References _confusion_matrix, ConfusionMatrix::calculate(), getModel(), getSampler(), and ConfusionMatrix::ready().

Referenced by main(), and RequestFile::makeModel().

Here is the call graph for this function:

EnvironmentPtr OpenModeller::getEnvironment ( )

inline

Returns the current environment related to model creation.

Returns: Pointer to environment related to model creation.

Definition at line 133 of file OpenModeller.hh.

Referenced by draw_occur(), and main().

AreaStats * OpenModeller::getEstimatedAreaStats ( double proportionAreaToSample = 0.01 )

Returns a pointer to the model AreaStats object which contains statistics about areas on the map generated by OM. This one uses only a random sample of the data points to estimate prediction areas. IMPORTANT: you should NOT delete the returned pointer!

Parameters

proportionAreaToSample Proportion of the area of interest (mask or intersection of all layers extents) to use as sample size.

Definition at line 413 of file OpenModeller.cpp.

References _env.

AreaStats * OpenModeller::getEstimatedAreaStats	(	const ConstEnvironmentPtr &	env,
		double	proportionAreaToSample = `0.01`
	)

Definition at line 421 of file OpenModeller.cpp.

References _alg, _estimatedAreaStats, AreaStats::addPrediction(), and AreaStats::reset().

Here is the call graph for this function:

Model OpenModeller::getModel ( ) const

inline

Returns model created by the algorithm.

Returns: Model object.

Definition at line 145 of file OpenModeller.hh.

Referenced by calculateModelStatistics(), getConfusionMatrix(), getRocCurve(), and main().

ConfigurationPtr OpenModeller::getModelConfiguration ( ) const

Definition at line 492 of file OpenModeller.cpp.

References _alg, _confusion_matrix, _roc_curve, _samp, RocCurve::getConfiguration(), ConfusionMatrix::getConfiguration(), RocCurve::ready(), and ConfusionMatrix::ready().

Referenced by main(), and RequestFile::makeModel().

Here is the call graph for this function:

RocCurve *const OpenModeller::getRocCurve ( )

Returns a pointer to the internal object representing the ROC curve that was calculated from the last generated model. IMPORTANT: you should NOT delete the returned pointer!

Definition at line 476 of file OpenModeller.cpp.

References _roc_curve, RocCurve::calculate(), getModel(), getSampler(), and RocCurve::ready().

Referenced by main(), and RequestFile::makeModel().

Here is the call graph for this function:

const SamplerPtr& OpenModeller::getSampler ( ) const

inline

Returns current sampler setting.

Returns: Pointer to sampler.

Definition at line 151 of file OpenModeller.hh.

Referenced by calculateModelStatistics(), getConfusionMatrix(), getRocCurve(), main(), and RequestFile::makeModel().

Scalar OpenModeller::getValue	(	const ConstEnvironmentPtr &	env,
		Coord	x,
		Coord	y
	)

Get prediction at a given point.

Parameters

env	Pointer to Environment class with the layers to get environmental values from.
x	X coordinate of point being queried
y	Y coordinate of point being queried

Returns: Prediction value at the specified point. Valid values range from 0.0 to 1.0. Value -1.0 means there is no prediction for that point (masked or not predicted)

Definition at line 369 of file OpenModeller.cpp.

References _alg, _env, and Sample::size().

Referenced by draw_niche().

Here is the call graph for this function:

Scalar OpenModeller::getValue ( Scalar const * environment_values )

Get prediction at a given point.

Parameters

environment_values Vector with environment values.

Returns: Prediction value at the specified point. Valid values range from 0.0 to 1.0. Value -1.0 means there is no prediction for that point (masked or not predicted)

Definition at line 395 of file OpenModeller.cpp.

References _alg, and _env.

std::string OpenModeller::getVersion ( )

Returns the openModeller client interface version in the format "n.m"

Definition at line 109 of file OpenModeller.cpp.

Referenced by main().

bool OpenModeller::hasEnvironment ( )

private

Definition at line 141 of file OpenModeller.cpp.

References _env, and _presence.

Referenced by setAlgorithm().

int OpenModeller::numAvailableAlgorithms ( )

Number of available algorithms. If the algorithms are not already searched in the system, searchs them first.

Returns: Number of available algorithms.

Definition at line 135 of file OpenModeller.cpp.

References AlgorithmFactory::numAvailableAlgorithms().

Here is the call graph for this function:

int OpenModeller::run ( )

inline

Compatibility with old oM client versions.

Definition at line 283 of file OpenModeller.hh.

void OpenModeller::setAbortionCallback	(	AbortionCallback	func,
		void *	param = `0`
	)

Sets a callback function to be called during model creation and map projection to check if the job should be aborted.

Parameters

func	Pointer to the callback function.
param	User parameter to be passed to the callback function.

Definition at line 65 of file OpenModeller.cpp.

References _callback_wrapper, and CallbackWrapper::setAbortionCallback().

Here is the call graph for this function:

int OpenModeller::setAlgorithm	(	std::string const	id,
		int	nparam,
		AlgParameter const *	param
	)

Define algorithm that will be used to generate the model.

Parameters

id	Algorithm's identifier. Must match the Algorithm::getID() returned string.
nparam	Number of parameters.
param	Vector with all parameters. The address 'param' points to must exist when the method "run()" is called.

Returns: zero if something goes wrong like the algorithm ID does not exist, use different number of parameters, etc.

Definition at line 185 of file OpenModeller.cpp.

References _absence, _alg, _env, _error, _presence, _samp, createSampler(), Log::error(), hasEnvironment(), Log::instance(), AlgorithmFactory::newAlgorithm(), setSampler(), and Log::warn().

Referenced by RequestFile::_setAlgorithm(), and main().

Here is the call graph for this function:

void OpenModeller::setEnvironment	(	std::vector< std::string >	categ_map,
		std::vector< std::string >	continuous_map,
		const std::string &	mask
	)

Defines environmental layers and the mask using STL arguments. Also creates the Environment object used for native range projection.

Parameters

categ_map	Vector of strings containing the file names of categorical map layers.
continuous_map	Vector of strings containing the file names of continuous map layers.
mask	File name of the mask map layer.

Definition at line 167 of file OpenModeller.cpp.

References _env, and createEnvironment().

Referenced by RequestFile::_setEnvironment().

Here is the call graph for this function: