GUIFeatures.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) 1999-2008 Soeren Sonnenburg
00008  * Written (W) 1999-2008 Gunnar Raetsch
00009  * Copyright (C) 1999-2008 Fraunhofer Institute FIRST and Max-Planck-Society
00010  */
00011 
00012 #ifndef __GUIFEATURES__H
00013 #define __GUIFEATURES__H
00014 
00015 #include <shogun/lib/config.h>
00016 #include <shogun/base/SGObject.h>
00017 #include <shogun/features/Labels.h>
00018 #include <shogun/features/Features.h>
00019 #include <shogun/features/RealFileFeatures.h>
00020 #include <shogun/features/TOPFeatures.h>
00021 #include <shogun/features/FKFeatures.h>
00022 #include <shogun/features/StringFeatures.h>
00023 #include <shogun/features/SimpleFeatures.h>
00024 #include <shogun/features/SparseFeatures.h>
00025 #include <shogun/features/CombinedFeatures.h>
00026 #include <shogun/features/CombinedDotFeatures.h>
00027 #include <shogun/features/WDFeatures.h>
00028 #include <shogun/features/ExplicitSpecFeatures.h>
00029 #include <shogun/features/ImplicitWeightedSpecFeatures.h>
00030 
00031 namespace shogun
00032 {
00033 class CSGInterface;
00034 
00036 class CGUIFeatures : public CSGObject
00037 {
00039     enum EFeatureType
00040     {
00041         Simple,
00042         Sparse
00043     };
00044 
00045     public:
00047         CGUIFeatures() {};
00051         CGUIFeatures(CSGInterface* interface);
00053         ~CGUIFeatures();
00054 
00056         inline CFeatures *get_train_features() { return train_features; }
00058         inline CFeatures *get_test_features() { return test_features; }
00059 
00063         inline bool set_train_features(CFeatures* f)
00064         { 
00065             //invalidate_train();
00066             SG_UNREF(train_features);
00067             SG_REF(f);
00068             train_features=f;
00069             return true;
00070         }
00071 
00075         inline bool set_test_features(CFeatures* f)
00076         { 
00077             //invalidate_test();
00078             SG_UNREF(test_features);
00079             SG_REF(f);
00080             test_features=f;
00081             return true;
00082         }
00083 
00087         void add_train_features(CFeatures* f);
00091         void add_test_features(CFeatures* f);
00095         void add_train_dotfeatures(CDotFeatures* f);
00099         void add_test_dotfeatures(CDotFeatures* f);
00100 
00102         bool del_last_feature_obj(char* target);
00103 
00105         void invalidate_train();
00107         void invalidate_test();
00108 
00110         bool load(
00111             char* filename, char* fclass, char* type, char* target,
00112             int32_t size, int32_t comp_features);
00114         bool save(char* filename, char* type, char* target);
00116         bool clean(char* target);
00118         bool reshape(char* target, int32_t num_feat, int32_t num_vec);
00119 
00121         CFeatures* get_convert_features(char* target);
00123         bool set_convert_features(CFeatures* features, char* target);
00124 
00128         CSparseFeatures<float64_t>* convert_simple_real_to_sparse_real(
00129             CSimpleFeatures<float64_t>* src);
00133         CStringFeatures<char>* convert_simple_char_to_string_char(
00134             CSimpleFeatures<char>* src);
00139         CSimpleFeatures<float64_t>* convert_simple_char_to_simple_align(
00140             CSimpleFeatures<char>* src,
00141             float64_t gap_cost=0);
00145         CSimpleFeatures<float64_t>* convert_simple_word_to_simple_salzberg(
00146             CSimpleFeatures<uint16_t>* src);
00147 
00151         CTOPFeatures* convert_string_word_to_simple_top(
00152             CStringFeatures<uint16_t>* src);
00156         CFKFeatures* convert_string_word_to_simple_fk(
00157             CStringFeatures<uint16_t>* src);
00161         CSimpleFeatures<float64_t>* convert_sparse_real_to_simple_real(
00162             CSparseFeatures<float64_t>* src);
00167         CExplicitSpecFeatures* convert_string_byte_to_spec_word(
00168                 CStringFeatures<uint16_t>* src, bool use_norm);
00169 
00178         template <class CT, class ST>
00179         CStringFeatures<ST>* convert_string_char_to_string_generic(
00180             CStringFeatures<CT>* src,
00181             int32_t order=1, int32_t start=0, int32_t gap=0, char rev='f', CAlphabet* alpha=NULL)
00182         {
00183             if (src && src->get_feature_class()==C_STRING)
00184             {
00185                 //create dense features with 0 cache
00186                 SG_INFO("Converting CT STRING features to ST STRING ones (order=%i).\n",order);
00187                 bool free_alpha=false;
00188                 
00189                 if (!alpha)
00190                 {
00191                     CAlphabet* a = src->get_alphabet();
00192 
00193                     if ( a && a->get_alphabet() == DNA )
00194                         alpha=new CAlphabet(RAWDNA);
00195                     else
00196                         alpha=new CAlphabet(a);
00197 
00198                     free_alpha=true;
00199                     SG_UNREF(a);
00200                 }
00201 
00202                 CStringFeatures<ST>* sf=new CStringFeatures<ST>(alpha);
00203                 if (sf && sf->obtain_from_char_features(src, start, order, gap, rev=='r'))
00204                 {
00205                     SG_INFO("Conversion was successful.\n");
00206                     return sf;
00207                 }
00208 
00209                 if (free_alpha)
00210                     SG_UNREF(alpha);
00211                 SG_UNREF(sf);
00212             }
00213             else
00214                 SG_ERROR("No features of class/type STRING/CT available.\n");
00215 
00216             return NULL;
00217         }
00218 
00219 
00221         bool set_reference_features(char* target);
00222 
00224         inline virtual const char* get_name() const { return "GUIFeatures"; }
00225 
00226     protected:
00228         CSGInterface* ui;
00230         CFeatures *train_features;
00232         CFeatures *test_features;
00234         CFeatures *ref_features;
00235 };
00236 }
00237 #endif
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines

SHOGUN Machine Learning Toolbox - Documentation