ppforest2::pp::PPStrategy Struct Reference

Abstract strategy for projection pursuit optimization. More...

#include <PPStrategy.hpp>

Public Member Functions
virtual types::Feature	index (types::FeatureMatrix const &x, stats::GroupPartition const &group_spec, Projector const &projector) const =0
	Evaluate the projection pursuit index for a given projector.
Projector	operator() (types::FeatureMatrix const &x, stats::GroupPartition const &group_spec) const
	Convenience operator: optimize and return the projector only.
virtual PPResult	optimize (types::FeatureMatrix const &x, stats::GroupPartition const &group_spec) const =0
	Find the optimal projection for the data.
Public Member Functions inherited from Strategy< PPStrategy >
virtual	~Strategy ()=default
virtual std::string	display_name () const=0
	Human-readable name for display in summaries.
virtual void	to_json (nlohmann::json &j) const=0
	Serialize this strategy's configuration to JSON.

Additional Inherited Members
Public Types inherited from Strategy< PPStrategy >
using	Factory
	Factory function type for deserializing a strategy from JSON.
using	Ptr
Static Public Member Functions inherited from Strategy< PPStrategy >
static Ptr	from_json (nlohmann::json const &j)
	Construct a strategy from its JSON representation.
static bool	register_strategy (std::string const &name, Factory factory)
	Register a concrete strategy for JSON deserialization.

Detailed Description

Abstract strategy for projection pursuit optimization.

Defines how to evaluate a projection (via an index function) and how to find the optimal projection for a given dataset and group partition.

Implementations must also provide to_json() for serialization.

Member Function Documentation

index()

virtual types::Feature ppforest2::pp::PPStrategy::index ( types::FeatureMatrix const & x, stats::GroupPartition const & group_spec, Projector const & projector ) const

pure virtual

Evaluate the projection pursuit index for a given projector.

Parameters

x	Feature matrix (n × p).
group_spec	Group partition.
projector	Projection vector (p).

Returns

Index value (higher is better separation).

Implemented in ppforest2::pp::PPPDAStrategy.

operator()()

Projector ppforest2::pp::PPStrategy::operator() ( types::FeatureMatrix const & x, stats::GroupPartition const & group_spec ) const

inline

Convenience operator: optimize and return the projector only.

optimize()

virtual PPResult ppforest2::pp::PPStrategy::optimize ( types::FeatureMatrix const & x, stats::GroupPartition const & group_spec ) const

pure virtual

Find the optimal projection for the data.

Parameters

x	Feature matrix (n × p).
group_spec	Group partition.

Returns

Optimized projector and its index value.

Implemented in ppforest2::pp::PPPDAStrategy.

---

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

• core/src/models/PPStrategy.hpp