Public Member Functions

CPreprocessor Class Reference


Detailed Description

Class Preprocessor defines a preprocessor interface.

Preprocessors are transformation functions that don't change the domain of the input features. These functions can be applied in-place if the input features fit in memory or can be applied on-the-fly when (depending on features) a feature caching strategy is applied. However, if the individual features are in $\bf{R}$ they have to stay in $\bf{R}$ although the dimensionality of the feature vectors is allowed change.

As preprocessors might need a certain initialization they may expect that the init() function is called before anything else. The actual preprocessing is feature type dependent and thus coordinated in the sub-classes, cf. e.g. CSimplePreprocessor .

Definition at line 74 of file Preprocessor.h.

Inheritance diagram for CPreprocessor:
Inheritance graph
[legend]

List of all members.

Public Member Functions

 CPreprocessor ()
virtual ~CPreprocessor ()
virtual bool init (CFeatures *f)=0
 initialize preprocessor from features
virtual void cleanup ()=0
 cleanup
virtual EFeatureType get_feature_type ()=0
virtual EFeatureClass get_feature_class ()=0
virtual EPreprocessorType get_type () const =0
 return a type of preprocessor

Constructor & Destructor Documentation

CPreprocessor (  ) 

constructor

Definition at line 17 of file Preprocessor.cpp.

~CPreprocessor (  )  [virtual]

destructor

Definition at line 22 of file Preprocessor.cpp.


Member Function Documentation

virtual void cleanup (  )  [pure virtual]
virtual EFeatureClass get_feature_class (  )  [pure virtual]
virtual EFeatureType get_feature_type (  )  [pure virtual]
virtual EPreprocessorType get_type (  )  const [pure virtual]
virtual bool init ( CFeatures f  )  [pure virtual]

The documentation for this class was generated from the following files:
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines

SHOGUN Machine Learning Toolbox - Documentation