ppforest2::dr::DRStrategy Struct Reference

Abstract strategy for dimensionality reduction. More...

#include <DRStrategy.hpp>

Public Member Functions
DRResult	operator() (types::FeatureMatrix const &x, stats::GroupPartition const &group_spec, stats::RNG &rng) const
	Convenience operator: equivalent to select().
virtual DRResult	select (types::FeatureMatrix const &x, stats::GroupPartition const &group_spec, stats::RNG &rng) const =0
	Select a subset of variables.
Public Member Functions inherited from Strategy< DRStrategy >
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< DRStrategy >
using	Factory
	Factory function type for deserializing a strategy from JSON.
using	Ptr
Static Public Member Functions inherited from Strategy< DRStrategy >
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 dimensionality reduction.

Before projection pursuit optimization, a DR strategy selects a subset of variables (columns) to work with. This reduces the cost of the PP step and introduces randomness in forests.

Implementations must also provide to_json() for serialization.

Member Function Documentation

operator()()

DRResult ppforest2::dr::DRStrategy::operator() ( types::FeatureMatrix const & x, stats::GroupPartition const & group_spec, stats::RNG & rng ) const

inline

Convenience operator: equivalent to select().

select()

virtual DRResult ppforest2::dr::DRStrategy::select ( types::FeatureMatrix const & x, stats::GroupPartition const & group_spec, stats::RNG & rng ) const

pure virtual

Select a subset of variables.

Parameters

x	Feature matrix (n × p).
group_spec	Group partition.
rng	Random number generator.

Returns

DRResult recording which columns were selected.

Implemented in ppforest2::dr::DRNoopStrategy, and ppforest2::dr::DRUniformStrategy.

---

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

• core/src/models/DRStrategy.hpp