WDFeatures.h

Go to the documentation of this file.
00001 /*
00002  * This program is free software; you can redistribute it and/or modify
00003  * it under the terms of the GNU General Public License as published by
00004  * the Free Software Foundation; either version 3 of the License, or
00005  * (at your option) any later version.
00006  *
00007  * Written (W) 2009-2010 Soeren Sonnenburg
00008  * Copyright (C) 2009 Fraunhofer Institute FIRST and Max-Planck-Society
00009  * Copyright (C) 2010 Berlin Institute of Technology
00010  */
00011 
00012 #ifndef _WDFEATURES_H___
00013 #define _WDFEATURES_H___
00014 
00015 #include <shogun/lib/common.h>
00016 #include <shogun/features/DotFeatures.h>
00017 #include <shogun/features/StringFeatures.h>
00018 
00019 namespace shogun
00020 {
00021 template <class ST> class CStringFeatures;
00022 
00028 class CWDFeatures : public CDotFeatures
00029 {
00030     public:
00032         CWDFeatures(void);
00033 
00040         CWDFeatures(CStringFeatures<uint8_t>* str, int32_t order, int32_t from_order);
00041 
00043         CWDFeatures(const CWDFeatures & orig);
00044 
00046         virtual ~CWDFeatures();
00047 
00055         inline virtual int32_t get_dim_feature_space() const
00056         {
00057             return w_dim;
00058         }
00059 
00067         virtual float64_t dot(int32_t vec_idx1, CDotFeatures* df, int32_t vec_idx2);
00068 
00075         virtual float64_t dense_dot(int32_t vec_idx1, const float64_t* vec2, int32_t vec2_len);
00076 
00085         virtual void add_to_dense_vec(float64_t alpha, int32_t vec_idx1, float64_t* vec2, int32_t vec2_len, bool abs_val=false);
00086 
00092         virtual inline int32_t get_nnz_features_for_vector(int32_t num)
00093         {
00094             int32_t vlen=-1;
00095             bool free_vec;
00096             uint8_t* vec=strings->get_feature_vector(num, vlen, free_vec);
00097             strings->free_feature_vector(vec, num, free_vec);
00098             return degree*vlen;
00099         }
00100 
00101         #ifndef DOXYGEN_SHOULD_SKIP_THIS
00102 
00103         struct wd_feature_iterator
00104         {
00106             uint8_t* vec;
00108             int32_t vidx;
00110             int32_t vlen;
00112             bool vfree;
00113 
00118             int32_t lim;
00119             int32_t* val;
00120             int32_t asize;
00121             int32_t asizem1;
00122             int32_t offs;
00123             int32_t k;
00124             int32_t i;
00125             int32_t o;
00127         };
00128         #endif
00129 
00139         virtual void* get_feature_iterator(int32_t vector_index);
00140 
00151         virtual bool get_next_feature(int32_t& index, float64_t& value, void* iterator);
00152 
00158         virtual void free_feature_iterator(void* iterator);
00159 
00164         virtual CFeatures* duplicate() const;
00165 
00170         inline virtual EFeatureType get_feature_type()
00171         {
00172             return F_UNKNOWN;
00173         }
00174 
00179         inline virtual EFeatureClass get_feature_class()
00180         {
00181             return C_WD;
00182         }
00183 
00184         inline virtual int32_t get_num_vectors() const
00185         {
00186             return num_strings;
00187         }
00188 
00189         inline virtual int32_t get_size()
00190         {
00191             return sizeof(float64_t);
00192         }
00193 
00196         void set_normalization_const(float64_t n=0);
00197 
00199         inline float64_t get_normalization_const()
00200         {
00201             return normalization_const;
00202         }
00203 
00205         inline virtual const char* get_name() const { return "WDFeatures"; }
00206 
00213         void set_wd_weights(float64_t* weights, int32_t d)
00214         {
00215             ASSERT(d==degree);
00216 
00217             for (int32_t i=0; i<degree; i++)
00218                 wd_weights[i]=weights[i];
00219         }
00220 
00222         void set_wd_weights();
00223 
00224     protected:
00226         CStringFeatures<uint8_t>* strings;
00227 
00229         int32_t degree;
00231         int32_t from_degree;
00233         int32_t string_length;
00235         int32_t num_strings;
00237         int32_t alphabet_size;
00239         int32_t w_dim;
00241         float64_t* wd_weights;
00242 
00244         float64_t normalization_const;
00245 
00246 };
00247 }
00248 #endif // _WDFEATURES_H___
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines

SHOGUN Machine Learning Toolbox - Documentation