FMT 1.2.0
Forest management tools for forest planning
Loading...
Searching...
No Matches
Public Member Functions | Static Public Member Functions | List of all members
Core::FMTYieldModelPools Class Reference

Machine learning yield model predicting carbon pools. More...

#include <FMTYieldModelPools.h>

Inheritance diagram for Core::FMTYieldModelPools:
[legend]
Collaboration diagram for Core::FMTYieldModelPools:
[legend]

Public Member Functions

 FMTYieldModelPools ()=default
 Default constructor for FMTYieldModelPools.
 
 FMTYieldModelPools (const FMTYieldModelPools &rhs)=default
 Copy constructor for FMTYieldModelPools.
 
FMTYieldModelPoolsoperator= (const FMTYieldModelPools &rhs)=default
 Copy assignment operator for FMTYieldModelPools.
 
 ~FMTYieldModelPools ()=default
 Default destructor for FMTYieldModelPools.
 
 FMTYieldModelPools (const boost::property_tree::ptree &jsonProps, std::vector< std::string > &inputYields)
 Construct a FMTYieldModelPools from a JSON tree and an input yield name list.
 
std::unique_ptr< FMTYieldModelclone () const
 clone the yield model.
 
const std::vector< double > getInputValues (const Graph::FMTPredictor &predictor) const
 Return the input values for a predictor.
 
- Public Member Functions inherited from Core::FMTYieldModelNn
virtual ~FMTYieldModelNn ()
 Destructor for FMTYieldModelNn.
 
 FMTYieldModelNn ()=default
 Default constructor for FMTYieldModelNn.
 
 FMTYieldModelNn (const FMTYieldModelNn &rhs)
 Copy constructor for FMTYieldModelNn.
 
const std::vector< double > predict (const Core::FMTYieldRequest &request) const
 Run the machine learning model to predict its outputs for a request.
 
- Public Member Functions inherited from Core::FMTYieldModel
virtual void setModel (Models::FMTModel *p_modelPtr)
 Set the model linked to the yield model.
 
virtual ~FMTYieldModel ()
 Destructor for FMTYieldModel.
 
 FMTYieldModel ()
 Constructor for FMTYieldModel.
 
 FMTYieldModel (const FMTYieldModel &rhs)=default
 Copy constructor for FMTYieldModel.
 
FMTYieldModeloperator= (const FMTYieldModel &rhs)=default
 Default copy assignment operator for FMTYieldModel.
 
const std::string & getModelName () const
 Return the model name.
 
const std::vector< std::string > & getModelYields () const
 Return the model yield names.
 
virtual std::unique_ptr< FMTYieldModelclone () const =0
 clone the yield model.
 
virtual const std::vector< double > predict (const Core::FMTYieldRequest &request) const =0
 Predict the yield for a request.
 
virtual std::unique_ptr< FMTYieldModelpreSolve (const FMTMaskFilter &filter, const std::vector< FMTTheme > &newthemes) const
 Return a presolved copy of the yield model, or the same model by default.
 
virtual std::unique_ptr< FMTYieldModelpostSolve (const FMTMaskFilter &filter, const std::vector< FMTTheme > &basethemes) const
 Return a postsolved copy of the yield model, a clone by default.
 
virtual operator std::string () const
 Return the string representation of the yield model.
 
virtual std::vector< std::vector< double > > getPeriodicValues () const
 Try to turn the yield model into periodic constant values.
 
virtual void clearRandomYieldsCache ()
 Clear the cache of all random yield models.
 
- Public Member Functions inherited from Core::FMTObject
 FMTObject ()
 Default constructor for FMTObject.
 
virtual ~FMTObject ()
 Default virtual destructor for FMTObject.
 
 FMTObject (const std::unique_ptr< Exception::FMTExceptionHandler > exhandler)
 Construct a FMTObject passing in the exception handler of another FMTObject.
 
 FMTObject (const FMTObject &rhs)
 Copy constructor for FMTObject.
 
FMTObjectoperator= (const FMTObject &rhs)
 Copy assignment operator for FMTObject.
 
virtual void passInLogger (const std::unique_ptr< Logging::FMTLogger > &logger)
 Pass in the logger of another FMTObject.
 
void passInExceptionHandler (const std::unique_ptr< Exception::FMTExceptionHandler > &exhandler)
 Pass in the exception handler of another FMTObject.
 
void redirectLogToFile (const std::string &location)
 Redirect the log to a specific file, appending to it.
 
virtual void setDefaultLogger ()
 Create and set a default logger to the FMTObject.
 
virtual void setQuietLogger ()
 Create and set a quiet logger to the FMTObject.
 
virtual void setTaskLogger ()
 Create and set a task logger to the FMTObject.
 
virtual void setDebugLogger ()
 Create and set a debug logger to the FMTObject.
 
void setDefaultExceptionHandler ()
 Create and set a default exception handler to the FMTObject.
 
void setQuietExceptionHandler ()
 Create and set a quiet exception handler to the FMTObject.
 
void setDebugExceptionHandler ()
 Create and set a debug exception handler to the FMTObject.
 
void setFreeExceptionHandler ()
 Create and set a free exception handler to the FMTObject.
 
void disableNestedExceptions ()
 Disable the nested exception throw of the exception handler.
 
void enableNestedExceptions ()
 Enable the nested exception throw of the exception handler.
 
void setErrorsToWarnings (const std::vector< Exception::FMTexc > &errors)
 Set a list of errors to be cast to warnings on the exception handler.
 
void setMaxWarningsBeforeSilenced (const size_t &maxwarningcount)
 Set the number of warnings raised before being silenced.
 
template<class Archive >
void serialize (Archive &ar, const unsigned int version)
 

Static Public Member Functions

static std::string getModelType ()
 Return the model type of the yield model.
 
- Static Public Member Functions inherited from Core::FMTObject
static Logging::FMTLoggergetLogger ()
 get a pointer to the actual logger.
 
static Exception::FMTExceptionHandlergetExceptionHandler ()
 get a pointer to the actual exception handler.
 
static std::string getRuntimeLocation ()
 Return the location of the FMT shared library.
 
static unsigned long long getAvailableMemory ()
 Return the available memory in bytes.
 
static void setTerminateStack ()
 will write the stack in the log when terminate called and raise a function failed.
 
static void setAbortStack ()
 will write the stack in the log when abort called with SIGABRT and raise a function failed.
 

Additional Inherited Members

- Protected Member Functions inherited from Core::FMTYieldModelNn
void _validateInputYields (std::vector< std::string > &expectedYields, std::vector< std::string > &inputYields) const
 Validate that there is the expected number of inputs in the model.
 
const std::string & getModelType () const
 Return the model type.
 
const std::vector< float > & _getStandardParamMeans () const
 Return the input variable means used in the standardization process.
 
const std::vector< float > & _getStandardParamVars () const
 Return the input variable variances used in the standardization process.
 
const std::vector< std::string > & getModelOutputNames () const
 Return the model output names.
 
virtual const std::vector< double > getInputValues (const Graph::FMTPredictor &predictor) const =0
 Return the input values based on a predictor.
 
const void removeNans (std::vector< float > &input) const
 Replace nan values with default values.
 
 FMTYieldModelNn (const boost::property_tree::ptree &jsonProps, std::vector< std::string > &inputYields)
 Construct a FMTYieldModelNn from a JSON tree and an input yield name list.
 
- Protected Member Functions inherited from Core::FMTObject
template<class Archive >
void forceSave (Archive &ar, const unsigned int version) const
 Force the serialization to save useful information, which the default FMTObject serialization does not.
 
template<class Archive >
void forceLoad (Archive &ar, const unsigned int version)
 Force the serialization to load useful information, which the default FMTObject serialization does not.
 
void checkSignals () const
 Check if the user has sent a ctrl-c signal using boost::python to FMT.
 
void setCplHandler ()
 Pass the FMT exception handler to the GDAL exception handler (GDAL only).
 
- Static Protected Member Functions inherited from Core::FMTYieldModelNn
static const std::vector< std::string > _getNextLineAndSplitIntoTokens (std::istream &str)
 Read a CSV file line by line, splitting into tokens.
 
static const std::vector< float > _standardize (std::vector< float > &input, const std::vector< float > &means, const std::vector< float > &vars)
 Apply the standardization feature scaling to the inputs of a machine learning model.
 
- Static Protected Member Functions inherited from Core::FMTObject
static std::chrono::time_point< std::chrono::high_resolution_clock > getClock ()
 Return a clock of the current time.
 
template<class chrono >
static double getDuration (const std::chrono::time_point< std::chrono::high_resolution_clock > &startclock)
 Return the time elapsed since a start clock as a double.
 
static std::string getDurationInSeconds (const std::chrono::time_point< std::chrono::high_resolution_clock > &startclock)
 Return the time elapsed since a start clock in seconds as a string.
 
static void _logStack ()
 Log the stack trace...
 
static void _terminate ()
 Raise an error with the boost stacktrace.
 
static void _abort (int p_signal)
 Raise an error with the boost stacktrace on abort.
 
- Protected Attributes inherited from Core::FMTYieldModelNn
std::unique_ptr< Ort::Session > m_sessionPtr
 
const std::string m_JSON_PROP_MODEL_TYPE = "modelType"
 
const std::string m_JSON_PROP_MODEL_YIELDS = "modelYields"
 
const std::string m_JSON_PROP_MODEL_OUTPUTS = "outputNames"
 
const std::string m_JSON_PROP_STAND_FILE_PATH = "csvStandardisationFile"
 
std::string m_modelType
 
std::vector< float > m_standardParamMeans = {}
 
std::vector< float > m_standardParamVars = {}
 
std::vector< std::string > m_modelOutputs = {}
 
- Protected Attributes inherited from Core::FMTYieldModel
std::string m_modelName
 
std::vector< std::string > m_modelYields = {}
 
const std::string m_JSON_PROP_MODEL_NAME = "modelFileName"
 
Models::FMTModelm_modelPtr =nullptr
 
- Static Protected Attributes inherited from Core::FMTYieldModelNn
static std::unique_ptr< Ort::Env > m_envPtr
 
static const float m_UNKNOWN_DISTURBANCE_CODE
 
- Static Protected Attributes inherited from Core::FMTObject
static std::unique_ptr< Exception::FMTExceptionHandler_exhandler
 A shared pointer to the exception handler.
 
static std::unique_ptr< Logging::FMTLogger_logger
 A shared pointer to the logger.
 

Detailed Description

Machine learning yield model predicting carbon pools.

Constructor & Destructor Documentation

◆ FMTYieldModelPools() [1/3]

Core::FMTYieldModelPools::FMTYieldModelPools ( )
default

Default constructor for FMTYieldModelPools.

◆ FMTYieldModelPools() [2/3]

Core::FMTYieldModelPools::FMTYieldModelPools ( const FMTYieldModelPools rhs)
default

Copy constructor for FMTYieldModelPools.

Parameters
[in]rhsthe FMTYieldModelPools to copy.

◆ ~FMTYieldModelPools()

Core::FMTYieldModelPools::~FMTYieldModelPools ( )
default

Default destructor for FMTYieldModelPools.

◆ FMTYieldModelPools() [3/3]

Core::FMTYieldModelPools::FMTYieldModelPools ( const boost::property_tree::ptree jsonProps,
std::vector< std::string > &  inputYields 
)

Construct a FMTYieldModelPools from a JSON tree and an input yield name list.

Parameters
[in]jsonPropsthe JSON properties.
[in,out]inputYieldsthe input yield names.

Member Function Documentation

◆ clone()

std::unique_ptr< FMTYieldModel > Core::FMTYieldModelPools::clone ( ) const
virtual

clone the yield model.

Returns
a unique pointer to the cloned yield model.

Implements Core::FMTYieldModel.

◆ getInputValues()

const std::vector< double > Core::FMTYieldModelPools::getInputValues ( const Graph::FMTPredictor predictor) const
virtual

Return the input values for a predictor.

Parameters
[in]predictorthe predictor.
Returns
the input values.

Implements Core::FMTYieldModelNn.

◆ getModelType()

static std::string Core::FMTYieldModelPools::getModelType ( )
static

Return the model type of the yield model.

Returns
the model type.

◆ operator=()

FMTYieldModelPools & Core::FMTYieldModelPools::operator= ( const FMTYieldModelPools rhs)
default

Copy assignment operator for FMTYieldModelPools.

Parameters
[in]rhsthe FMTYieldModelPools to copy.
Returns
a reference to this FMTYieldModelPools.

The documentation for this class was generated from the following file: