forpy  2
forpy::IThreshOpt Class Reference

Find an optimal threshold. More...

#include <ithreshopt.h>

Inheritance diagram for forpy::IThreshOpt:
forpy::ClassificationOpt forpy::RegressionOpt forpy::FastClassOpt

Public Member Functions

virtual ~IThreshOpt ()
 
virtual std::shared_ptr< IThreshOptcreate_duplicate (const uint &random_seed) const VIRTUAL_PTR
 
virtual bool supports_weights () const
 Whether this threshold optimizer can take into account weights during the optimization. More...
 
virtual void check_annotations (IDataProvider *dprov) VIRTUAL_VOID
 Validate annotations for usability with this optimizer. More...
 
virtual void transfer_or_run_check (IThreshOpt *other, IDataProvider *dprov) VIRTUAL_VOID
 
virtual void full_entropy (const IDataProvider &dprov, Desk *) const VIRTUAL_VOID
 Get the full entropy for one node before optimization. More...
 
virtual void optimize (Desk *) const VIRTUAL_VOID
 Optimize for one node. More...
 
virtual float get_gain_threshold_for (const size_t &node_id) VIRTUAL(float)
 Get the gain threshold to use for this node. More...
 
virtual bool operator== (const IThreshOpt &rhs) const VIRTUAL(bool)
 

Protected Member Functions

 IThreshOpt ()
 

Private Member Functions

template<class Archive >
void serialize (Archive &, const uint &)
 

Friends

class cereal::access
 

Detailed Description

Find an optimal threshold.

This class is classically used by a forpy::IDecider to optimize the threshold ( \(\tau\)).

Definition at line 23 of file ithreshopt.h.

Constructor & Destructor Documentation

◆ ~IThreshOpt()

virtual forpy::IThreshOpt::~IThreshOpt ( )
inlinevirtual

Definition at line 25 of file ithreshopt.h.

◆ IThreshOpt()

forpy::IThreshOpt::IThreshOpt ( )
inlineprotected

For deserialization.

Definition at line 61 of file ithreshopt.h.

Member Function Documentation

◆ check_annotations()

virtual void forpy::IThreshOpt::check_annotations ( IDataProvider dprov)
virtual

Validate annotations for usability with this optimizer.

Reimplemented in forpy::ClassificationOpt, and forpy::RegressionOpt.

◆ create_duplicate()

virtual std::shared_ptr<IThreshOpt> forpy::IThreshOpt::create_duplicate ( const uint random_seed) const
virtual

◆ full_entropy()

virtual void forpy::IThreshOpt::full_entropy ( const IDataProvider dprov,
Desk  
) const
virtual

Get the full entropy for one node before optimization.

Setup tasks can be performed within this function. If the entropy determined is below a certain threshold, the optimization is stopped (see, e.g., forpy::FastDecider).

Reimplemented in forpy::ClassificationOpt, forpy::RegressionOpt, and forpy::FastClassOpt.

◆ get_gain_threshold_for()

virtual float forpy::IThreshOpt::get_gain_threshold_for ( const size_t &  node_id)
virtual

Get the gain threshold to use for this node.

Reimplemented in forpy::ClassificationOpt, and forpy::RegressionOpt.

◆ operator==()

virtual bool forpy::IThreshOpt::operator== ( const IThreshOpt rhs) const
virtual

◆ optimize()

virtual void forpy::IThreshOpt::optimize ( Desk ) const
virtual

Optimize for one node.

Reimplemented in forpy::ClassificationOpt, forpy::RegressionOpt, and forpy::FastClassOpt.

◆ serialize()

template<class Archive >
void forpy::IThreshOpt::serialize ( Archive &  ,
const uint  
)
inlineprivate

Definition at line 66 of file ithreshopt.h.

◆ supports_weights()

virtual bool forpy::IThreshOpt::supports_weights ( ) const
inlinevirtual

Whether this threshold optimizer can take into account weights during the optimization.

By default, return false.

Definition at line 34 of file ithreshopt.h.

◆ transfer_or_run_check()

virtual void forpy::IThreshOpt::transfer_or_run_check ( IThreshOpt other,
IDataProvider dprov 
)
virtual

Friends And Related Function Documentation

◆ cereal::access

friend class cereal::access
friend

Definition at line 61 of file ithreshopt.h.


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