PLearn 0.1
|
00001 // -*- C++ -*- 00002 00003 // RegressionTreeRegisters.h 00004 // Copyright (c) 1998-2002 Pascal Vincent 00005 // Copyright (C) 1999-2002 Yoshua Bengio and University of Montreal 00006 // Copyright (c) 2002 Jean-Sebastien Senecal, Xavier Saint-Mleux, Rejean Ducharme 00007 // 00008 // Redistribution and use in source and binary forms, with or without 00009 // modification, are permitted provided that the following conditions are met: 00010 // 00011 // 1. Redistributions of source code must retain the above copyright 00012 // notice, this list of conditions and the following disclaimer. 00013 // 00014 // 2. Redistributions in binary form must reproduce the above copyright 00015 // notice, this list of conditions and the following disclaimer in the 00016 // documentation and/or other materials provided with the distribution. 00017 // 00018 // 3. The name of the authors may not be used to endorse or promote 00019 // products derived from this software without specific prior written 00020 // permission. 00021 // 00022 // THIS SOFTWARE IS PROVIDED BY THE AUTHORS ``AS IS'' AND ANY EXPRESS OR 00023 // IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 00024 // OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN 00025 // NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, 00026 // SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED 00027 // TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR 00028 // PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF 00029 // LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING 00030 // NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS 00031 // SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 00032 // 00033 // This file is part of the PLearn library. For more information on the PLearn 00034 // library, go to the PLearn Web site at www.plearn.org 00035 00036 00037 /* ********************************************************************************* 00038 * $Id: RegressionTreeRegisters.h, v 1.0 2004/07/19 10:00:00 Bengio/Kegl/Godbout * 00039 * This file is part of the PLearn library. * 00040 ********************************************************************************* */ 00041 00042 #ifndef RegressionTreeRegisters_INC 00043 #define RegressionTreeRegisters_INC 00044 00045 #include <plearn/vmat/VMatrix.h> 00046 #include <plearn/base/stringutils.h> 00047 #include <plearn/math/TMat.h> 00048 #include <plearn/vmat/VMat.h> 00049 00052 #ifndef RTR_type 00053 #define RTR_type uint32_t 00054 #endif 00055 00056 #ifndef RTR_type_id 00057 #define RTR_type_id int16_t 00058 #endif 00059 00060 #ifndef RTR_target_t 00061 #define RTR_target_t real 00062 #endif 00063 00064 #ifndef RTR_weight_t 00065 #define RTR_weight_t real 00066 #endif 00067 00068 //if you are sur the their is no missing value in the training set 00069 //you can set its value to false for some speed up. 00070 #ifndef RTR_HAVE_MISSING 00071 #define RTR_HAVE_MISSING true 00072 #endif 00073 00074 namespace PLearn { 00075 using namespace std; 00076 00077 class RegressionTreeLeave; 00078 class RegressionTreeRegisters: public VMatrix 00079 { 00080 typedef VMatrix inherited; 00081 00082 private: 00083 00084 /* 00085 Build options: they have to be set before training 00086 */ 00087 00088 int report_progress; 00089 int verbosity; 00090 00091 /* 00092 Learnt options: they are sized and initialized if need be, at build() or reinitRegisters() 00093 */ 00094 00095 int next_id; 00096 00097 TMat<RTR_type> tsorted_row; 00098 TVec<RTR_type_id> leave_register; 00099 VMat tsource; 00100 Mat tsource_mat; 00101 //we put it in pair instead of two vector to speed up 00102 //the getAllRegisteredRow(leave_id, col, reg, target, weight, value) fct 00103 TVec<pair<RTR_target_t,RTR_weight_t> > target_weight; 00104 VMat source; 00105 00106 bool do_sort_rows; 00107 bool mem_tsource; 00108 bool have_missing; 00109 00110 mutable vector<bool> compact_reg; 00111 mutable int compact_reg_leave; 00112 00114 mutable PP<RegressionTreeLeave> tmp_leave; 00116 mutable Vec tmp_vec; 00117 00118 public: 00119 00120 RegressionTreeRegisters(); 00121 RegressionTreeRegisters(VMat source_, bool report_progress_ = false, 00122 bool vebosity_ = false, bool do_sort_rows = true, 00123 bool mem_tsource_ = true); 00124 RegressionTreeRegisters(VMat source_, TMat<RTR_type> tsorted_row_, 00125 VMat tsource_, bool report_progress_ = false, 00126 bool vebosity_ = false, bool do_sort_rows = true, 00127 bool mem_tsource_ = true); 00128 virtual ~RegressionTreeRegisters(); 00129 00130 PLEARN_DECLARE_OBJECT(RegressionTreeRegisters); 00131 00132 static void declareOptions(OptionList& ol); 00133 virtual void makeDeepCopyFromShallowCopy(CopiesMap &copies); 00134 virtual void build(); 00135 void reinitRegisters(); 00136 inline void registerLeave(RTR_type_id leave_id, int row) 00137 { leave_register[row] = leave_id; } 00138 inline virtual real get(int i, int j) const{ 00139 if(j<inputsize())return tsource->get(j,i); 00140 if(j==inputsize())return target_weight[i].first; 00141 else return target_weight[i].second; 00142 } 00143 inline real getTarget(int row)const 00144 {return target_weight[row].first;} 00145 inline real getWeight(int row)const{ 00146 return target_weight[row].second; 00147 } 00148 inline void setWeight(int row,real val){ 00149 target_weight[row].second = val; 00150 } 00151 inline bool haveMissing()const{return have_missing;} 00152 inline RTR_type_id getNextId(){ 00153 PLCHECK(next_id<std::numeric_limits<RTR_type_id>::max()); 00154 next_id += 1;return next_id;} 00155 void getAllRegisteredRow(RTR_type_id leave_id, TVec<RTR_type> ®)const; 00156 void getAllRegisteredRow(RTR_type_id leave_id, int col, TVec<RTR_type> ®)const; 00157 void getAllRegisteredRow(RTR_type_id leave_id, int col, TVec<RTR_type> ®, 00158 TVec<pair<RTR_target_t,RTR_weight_t> > &t_w, Vec &value)const; 00159 void getAllRegisteredRowLeave( 00160 RTR_type_id leave_id, int col, TVec<RTR_type> ®, 00161 TVec<pair<RTR_target_t,RTR_weight_t> > &t_w, Vec &value, 00162 PP<RegressionTreeLeave> missing_leave, 00163 PP<RegressionTreeLeave> left_leave, 00164 PP<RegressionTreeLeave> right_leave, 00165 TVec<RTR_type> &candidate)const; 00166 tuple<real,real,int> bestSplitInRow( 00167 RTR_type_id leave_id, int col, TVec<RTR_type> ®, 00168 PP<RegressionTreeLeave> left_leave, 00169 PP<RegressionTreeLeave> right_leave, 00170 Vec left_error, Vec right_error)const; 00171 void printRegisters(); 00172 void getExample(int i, Vec& input, Vec& target, real& weight); 00173 inline virtual void put(int i, int j, real value) 00174 { 00175 PLASSERT(inputsize()>0&&targetsize()>0); 00176 if(j!=inputsize()+targetsize()) 00177 PLERROR("In RegressionTreeRegisters::put - implemented the put of " 00178 "the weightsize only"); 00179 setWeight(i,value); 00180 } 00181 00183 inline TMat<RTR_type> getTSortedRow(){return tsorted_row;} 00184 inline VMat getTSource(){return tsource;} 00185 virtual void finalize(){tsorted_row = TMat<RTR_type>();} 00186 00187 private: 00188 void build_(); 00189 void sortRows(); 00190 void sortEachDim(int dim); 00191 void verbose(string msg, int level); 00192 void checkMissing(); 00193 00194 }; 00195 00196 DECLARE_OBJECT_PTR(RegressionTreeRegisters); 00197 00198 } // end of namespace PLearn 00199 00200 #endif 00201 00202 00203 /* 00204 Local Variables: 00205 mode:c++ 00206 c-basic-offset:4 00207 c-file-style:"stroustrup" 00208 c-file-offsets:((innamespace . 0)(inline-open . 0)) 00209 indent-tabs-mode:nil 00210 fill-column:79 00211 End: 00212 */ 00213 // vim: filetype=cpp:expandtab:shiftwidth=4:tabstop=8:softtabstop=4:encoding=utf-8:textwidth=79 :