libshogun-dev 3.2.0-7.3build4 / usr / include / shogun / features / Features.h

/*
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 3 of the License, or
 * (at your option) any later version.
 *
 * Written (W) 1999-2009 Soeren Sonnenburg
 * Written (W) 1999-2008 Gunnar Raetsch
 * Written (W) 2011-2012 Heiko Strathmann
 * Copyright (C) 1999-2009 Fraunhofer Institute FIRST and Max-Planck-Society
 */

#ifndef _CFEATURES__H__
#define _CFEATURES__H__

#include <shogun/lib/common.h>
#include <shogun/io/File.h>
#include <shogun/base/SGObject.h>
#include <shogun/preprocessor/Preprocessor.h>
#include <shogun/features/FeatureTypes.h>
#include <shogun/features/SubsetStack.h>
#include <shogun/lib/List.h>
#include <shogun/lib/DynamicObjectArray.h>
#include <shogun/lib/DynamicArray.h>

namespace shogun
{
	class CFile;
	class CPreprocessor;
	class CKernel;
}

namespace shogun
{

/** @brief The class Features is the base class of all feature objects.
 *
 * It can be understood as a dense real valued feature matrix (with e.g.
 * columns as single feature vectors), a set of strings, graphs or any other
 * arbitrary collection of objects. As a result this class is kept very general
 * and implements only very weak interfaces to
 *
 * - duplicate the Feature object
 * - obtain the feature type (like F_DREAL, F_SHORT ...)
 * - obtain the feature class (like Simple dense matrices, sparse or strings)
 * - obtain the number of feature "vectors"
 *
 *   In addition it provides helpers to check e.g. for compability of feature objects.
 *
 *   Currently there are 3 general feature classes, which are CDenseFeatures
 *   (dense matrices), CSparseFeatures (sparse matrices), CStringFeatures (a
 *   set of strings) from which all the specific features like CDenseFeatures<float64_t>
 *   (dense real valued feature matrices) are derived.
 *
 *
 *   (Multiple) Subsets (of subsets) may be supported by inheriting classes.
 *   Sub-classes may want to overwrite the subset_changed_post() method which is
 *   called automatically after each subset change.
 *   A subset is put onto a stack using the add_subset() method. The last added
 *   subset may be removed via remove_subset()
 */
class CFeatures : public CSGObject
{
	public:
		/** constructor
		 *
		 * @param size cache size
		 */
		CFeatures(int32_t size=0);

		/** copy constructor */
		CFeatures(const CFeatures& orig);

		/** constructor
		 *
		 * @param loader File object via which data shall be loaded
		 */
		CFeatures(CFile* loader);

		/** duplicate feature object
		 *
		 * abstract base method
		 *
		 * @return feature object
		 */
		virtual CFeatures* duplicate() const=0;

		virtual ~CFeatures();

		/** get feature type
		 *
		 * abstract base method
		 *
		 * @return templated feature type
		 */
		virtual EFeatureType get_feature_type() const=0;

		/** get feature class
		 *
		 * abstract base method
		 *
		 * @return feature class like STRING, SIMPLE, SPARSE...
		 */
		virtual EFeatureClass get_feature_class() const=0;

		/** add preprocessor
		 *
		 * @param p preprocessor to set
		 */
		virtual void add_preprocessor(CPreprocessor* p);

		/** delete preprocessor from list
		 *
		 * @param num index of preprocessor in list
		 */
		virtual void del_preprocessor(int32_t num);

		/** get specified preprocessor
		 *
		 * @param num index of preprocessor in list
		 */
		CPreprocessor* get_preprocessor(int32_t num) const;

		/** set applied flag for preprocessor
		 *
		 * @param num index of preprocessor in list
		 */
		void set_preprocessed(int32_t num);

		/** get whether specified preprocessor was already applied
		 *
		 * @param num index of preprocessor in list
		 */
		bool is_preprocessed(int32_t num) const;

		/** get the number of applied preprocs
		 *
		 * @return number of applied preprocessors
		 */
		int32_t get_num_preprocessed() const;

		/** get number of preprocessors
		 *
		 * @return number of preprocessors
		 */
		int32_t get_num_preprocessors() const;

		/** clears all preprocs */
		void clean_preprocessors();

		/** print preprocessors */
		void list_preprocessors();

		/** get cache size
		 *
		 * @return cache size
		 */
		int32_t get_cache_size() const;

		/** get number of examples/vectors, possibly corresponding to the current subset
		 *
		 * abstract base method
		 *
		 * @return number of examples/vectors (possibly of subset, if implemented)
		 */
		virtual int32_t get_num_vectors() const=0;

		/** in case there is a feature matrix allow for reshaping
		 *
		 * NOT IMPLEMENTED!
		 *
		 * @param num_features new number of features
		 * @param num_vectors new number of vectors
		 * @return if reshaping was successful
		 */
		virtual bool reshape(int32_t num_features, int32_t num_vectors);

		/** list feature object */
		void list_feature_obj() const;

		/** load features from file
		 *
		 * @param loader File object via which data shall be loaded
		 */
		virtual void load(CFile* loader);

		/** save features to file
		 *
		 * @param writer File object via which data shall be saved
		 */
		virtual void save(CFile* writer);

		/** check feature compatibility
		 *
		 * @param f features to check for compatibility
		 * @return if features are compatible
		 */
		bool check_feature_compatibility(CFeatures* f) const;

		/** check if features have given property
		 *
		 * @param p feature property
		 * @return if features have given property
		 */
		bool has_property(EFeatureProperty p) const;

		/** set property
		 *
		 * @param p kernel property to set
		 */
		void set_property(EFeatureProperty p);

		/** unset property
		 *
		 * @param p kernel property to unset
		 */
		void unset_property(EFeatureProperty p);

		/** Takes a list of feature instances and returns a new instance being
		 * a concatenation of a copy of this instace's data and the given
		 * instancess data. Note that the feature types have to be equal.
		 *
		 * NOT IMPLEMENTED!
		 *
		 * @param others list of feature objects to append
		 * @return new feature object which contains copy of data of this
		 * instance and given ones
		 */
		virtual CFeatures* create_merged_copy(CList* others)
		{
			SG_ERROR("%s::create_merged_copy() is not yet implemented!\n")
			return NULL;
		}

		/** Convenience method for method with same name and list as parameter.
		 *
		 * NOT IMPLEMENTED!
		 *
		 * @param other feature object to append
		 * @return new feature object which contains copy of data of this
		 * instance and of given one
		 */
		virtual CFeatures* create_merged_copy(CFeatures* other)
		{
			SG_ERROR("%s::create_merged_copy() is not yet implemented!\n")
			return NULL;
		}

		/** adds a subset of indices on top of the current subsets (possibly
		 * subset o subset. Calls subset_changed_post() afterwards
		 *
		 * @param subset subset of indices to add
		 * */
		virtual void add_subset(SGVector<index_t> subset);

		/** removes that last added subset from subset stack, if existing
		 * Calls subset_changed_post() afterwards */
		virtual void remove_subset();

		/** removes all subsets
		 * Calls subset_changed_post() afterwards */
		virtual void remove_all_subsets();

		/** returns subset stack
		 *
		 * @return subset stack
		 */
		virtual CSubsetStack* get_subset_stack();

		/** method may be overwritten to update things that depend on subset */
		virtual void subset_changed_post() {}

		/** Creates a new CFeatures instance containing copies of the elements
		 * which are specified by the provided indices.
		 *
		 * This method is needed for a KernelMachine to store its model data.
		 * NOT IMPLEMENTED!
		 *
		 * @param indices indices of feature elements to copy
		 * @return new CFeatures instance with copies of feature data
		 */
		virtual CFeatures* copy_subset(SGVector<index_t> indices);

	private:
		void init();

	private:
		/** feature properties */
		uint64_t  properties;

		/** size of cache in MB */
		int32_t cache_size;

		/** list of preprocessors */
		CDynamicObjectArray* preproc;

		/** i'th entry is true if features were already preprocessed with preproc i */
		CDynamicArray<bool>* preprocessed;

	protected:
		/** subset used for index transformations */
		CSubsetStack* m_subset_stack;
};
}
#endif