/usr/include/vigra/random_forest/rf_decisionTree.hxx is in libvigraimpex-dev 1.10.0+dfsg-3ubuntu2.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 | /************************************************************************/
/* */
/* Copyright 2008-2009 by Ullrich Koethe and Rahul Nair */
/* */
/* This file is part of the VIGRA computer vision library. */
/* The VIGRA Website is */
/* http://hci.iwr.uni-heidelberg.de/vigra/ */
/* Please direct questions, bug reports, and contributions to */
/* ullrich.koethe@iwr.uni-heidelberg.de or */
/* vigra@informatik.uni-hamburg.de */
/* */
/* Permission is hereby granted, free of charge, to any person */
/* obtaining a copy of this software and associated documentation */
/* files (the "Software"), to deal in the Software without */
/* restriction, including without limitation the rights to use, */
/* copy, modify, merge, publish, distribute, sublicense, and/or */
/* sell copies of the Software, and to permit persons to whom the */
/* Software is furnished to do so, subject to the following */
/* conditions: */
/* */
/* The above copyright notice and this permission notice shall be */
/* included in all copies or substantial portions of the */
/* Software. */
/* */
/* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND */
/* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES */
/* OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND */
/* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT */
/* HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, */
/* WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING */
/* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR */
/* OTHER DEALINGS IN THE SOFTWARE. */
/* */
/************************************************************************/
#ifndef VIGRA_RANDOM_FOREST_DT_HXX
#define VIGRA_RANDOM_FOREST_DT_HXX
#include <algorithm>
#include <map>
#include <numeric>
#include "vigra/multi_array.hxx"
#include "vigra/mathutil.hxx"
#include "vigra/array_vector.hxx"
#include "vigra/sized_int.hxx"
#include "vigra/matrix.hxx"
#include "vigra/random.hxx"
#include "vigra/functorexpression.hxx"
#include <vector>
#include "rf_common.hxx"
#include "rf_visitors.hxx"
#include "rf_nodeproxy.hxx"
namespace vigra
{
namespace detail
{
// todo FINALLY DECIDE TO USE CAMEL CASE OR UNDERSCORES !!!!!!
/* decisiontree classifier.
*
* This class is actually meant to be used in conjunction with the
* Random Forest Classifier
* - My suggestion would be to use the RandomForest classifier with
* following parameters instead of directly using this
* class (Preprocessing default values etc is handled in there):
*
* \code
* RandomForest decisionTree(RF_Traits::Options_t()
* .features_per_node(RF_ALL)
* .tree_count(1) );
* \endcode
*
* \todo remove the classCount and featurecount from the topology
* array. Pass ext_param_ to the nodes!
* \todo Use relative addressing of nodes?
*/
class DecisionTree
{
/* \todo make private?*/
public:
/* value type of container array. use whenever referencing it
*/
typedef Int32 TreeInt;
ArrayVector<TreeInt> topology_;
ArrayVector<double> parameters_;
ProblemSpec<> ext_param_;
unsigned int classCount_;
public:
/* \brief Create tree with parameters */
template<class T>
DecisionTree(ProblemSpec<T> ext_param)
:
ext_param_(ext_param),
classCount_(ext_param.class_count_)
{}
/* clears all memory used.
*/
void reset(unsigned int classCount = 0)
{
if(classCount)
classCount_ = classCount;
topology_.clear();
parameters_.clear();
}
/* learn a Tree
*
* \tparam StackEntry_t The Stackentry containing Node/StackEntry_t
* Information used during learning. Each Split functor has a
* Stack entry associated with it (Split_t::StackEntry_t)
* \sa RandomForest::learn()
*/
template < class U, class C,
class U2, class C2,
class StackEntry_t,
class Stop_t,
class Split_t,
class Visitor_t,
class Random_t >
void learn( MultiArrayView<2, U, C> const & features,
MultiArrayView<2, U2, C2> const & labels,
StackEntry_t const & stack_entry,
Split_t split,
Stop_t stop,
Visitor_t & visitor,
Random_t & randint);
template < class U, class C,
class U2, class C2,
class StackEntry_t,
class Stop_t,
class Split_t,
class Visitor_t,
class Random_t>
void continueLearn( MultiArrayView<2, U, C> const & features,
MultiArrayView<2, U2, C2> const & labels,
StackEntry_t const & stack_entry,
Split_t split,
Stop_t stop,
Visitor_t & visitor,
Random_t & randint,
//an index to which the last created exterior node will be moved (because it is not used anymore)
int garbaged_child=-1);
/* is a node a Leaf Node? */
inline bool isLeafNode(TreeInt in) const
{
return (in & LeafNodeTag) == LeafNodeTag;
}
/* data driven traversal from root to leaf
*
* traverse through tree with data given in features. Use Visitors to
* collect statistics along the way.
*/
template<class U, class C, class Visitor_t>
TreeInt getToLeaf(MultiArrayView<2, U, C> const & features,
Visitor_t & visitor) const
{
TreeInt index = 2;
while(!isLeafNode(topology_[index]))
{
visitor.visit_internal_node(*this, index, topology_[index],features);
switch(topology_[index])
{
case i_ThresholdNode:
{
Node<i_ThresholdNode>
node(topology_, parameters_, index);
index = node.next(features);
break;
}
case i_HyperplaneNode:
{
Node<i_HyperplaneNode>
node(topology_, parameters_, index);
index = node.next(features);
break;
}
case i_HypersphereNode:
{
Node<i_HypersphereNode>
node(topology_, parameters_, index);
index = node.next(features);
break;
}
#if 0
// for quick prototyping! has to be implemented.
case i_VirtualNode:
{
Node<i_VirtualNode>
node(topology_, parameters, index);
index = node.next(features);
}
#endif
default:
vigra_fail("DecisionTree::getToLeaf():"
"encountered unknown internal Node Type");
}
}
visitor.visit_external_node(*this, index, topology_[index],features);
return index;
}
/* traverse tree to get statistics
*
* Tree is traversed in order the Nodes are in memory (i.e. if no
* relearning//pruning scheme is utilized this will be pre order)
*/
template<class Visitor_t>
void traverse_mem_order(Visitor_t visitor) const
{
TreeInt index = 2;
Int32 ii = 0;
while(index < topology_.size())
{
if(isLeafNode(topology_[index]))
{
visitor
.visit_external_node(*this, index, topology_[index]);
}
else
{
visitor
._internal_node(*this, index, topology_[index]);
}
}
}
template<class Visitor_t>
void traverse_post_order(Visitor_t visitor, TreeInt start = 2) const
{
typedef TinyVector<double, 2> Entry;
std::vector<Entry > stack;
std::vector<double> result_stack;
stack.push_back(Entry(2, 0));
int addr;
while(!stack.empty())
{
addr = stack.back()[0];
NodeBase node(topology_, parameters_, stack.back()[0]);
if(stack.back()[1] == 1)
{
stack.pop_back();
double leftRes = result_stack.back();
double rightRes = result_stack.back();
result_stack.pop_back();
result_stack.pop_back();
result_stack.push_back(rightRes+ leftRes);
visitor.visit_internal_node(*this,
addr,
node.typeID(),
rightRes+leftRes);
}
else
{
if(isLeafNode(node.typeID()))
{
visitor.visit_external_node(*this,
addr,
node.typeID(),
node.weights());
stack.pop_back();
result_stack.push_back(node.weights());
}
else
{
stack.back()[1] = 1;
stack.push_back(Entry(node.child(0), 0));
stack.push_back(Entry(node.child(1), 0));
}
}
}
}
/* same thing as above, without any visitors */
template<class U, class C>
TreeInt getToLeaf(MultiArrayView<2, U, C> const & features) const
{
::vigra::rf::visitors::StopVisiting stop;
return getToLeaf(features, stop);
}
template <class U, class C>
ArrayVector<double>::iterator
predict(MultiArrayView<2, U, C> const & features) const
{
TreeInt nodeindex = getToLeaf(features);
switch(topology_[nodeindex])
{
case e_ConstProbNode:
return Node<e_ConstProbNode>(topology_,
parameters_,
nodeindex).prob_begin();
break;
#if 0
//first make the Logistic regression stuff...
case e_LogRegProbNode:
return Node<e_LogRegProbNode>(topology_,
parameters_,
nodeindex).prob_begin();
#endif
default:
vigra_fail("DecisionTree::predict() :"
" encountered unknown external Node Type");
}
return ArrayVector<double>::iterator();
}
template <class U, class C>
Int32 predictLabel(MultiArrayView<2, U, C> const & features) const
{
ArrayVector<double>::const_iterator weights = predict(features);
return argMax(weights, weights+classCount_) - weights;
}
};
template < class U, class C,
class U2, class C2,
class StackEntry_t,
class Stop_t,
class Split_t,
class Visitor_t,
class Random_t>
void DecisionTree::learn( MultiArrayView<2, U, C> const & features,
MultiArrayView<2, U2, C2> const & labels,
StackEntry_t const & stack_entry,
Split_t split,
Stop_t stop,
Visitor_t & visitor,
Random_t & randint)
{
this->reset();
topology_.reserve(256);
parameters_.reserve(256);
topology_.push_back(features.shape(1));
topology_.push_back(classCount_);
continueLearn(features,labels,stack_entry,split,stop,visitor,randint);
}
template < class U, class C,
class U2, class C2,
class StackEntry_t,
class Stop_t,
class Split_t,
class Visitor_t,
class Random_t>
void DecisionTree::continueLearn( MultiArrayView<2, U, C> const & features,
MultiArrayView<2, U2, C2> const & labels,
StackEntry_t const & stack_entry,
Split_t split,
Stop_t stop,
Visitor_t & visitor,
Random_t & randint,
//an index to which the last created exterior node will be moved (because it is not used anymore)
int garbaged_child)
{
std::vector<StackEntry_t> stack;
stack.reserve(128);
ArrayVector<StackEntry_t> child_stack_entry(2, stack_entry);
stack.push_back(stack_entry);
size_t last_node_pos = 0;
StackEntry_t top=stack.back();
while(!stack.empty())
{
// Take an element of the stack. Obvious ain't it?
top = stack.back();
stack.pop_back();
// Make sure no data from the last round has remained in Pipeline;
child_stack_entry[0].reset();
child_stack_entry[1].reset();
split.reset();
//Either the Stopping criterion decides that the split should
//produce a Terminal Node or the Split itself decides what
//kind of node to make
TreeInt NodeID;
if(stop(top))
NodeID = split.makeTerminalNode(features,
labels,
top,
randint);
else
{
//TIC;
NodeID = split.findBestSplit(features,
labels,
top,
child_stack_entry,
randint);
//std::cerr << TOC <<" " << NodeID << ";" <<std::endl;
}
// do some visiting yawn - just added this comment as eye candy
// (looks odd otherwise with my syntax highlighting....
visitor.visit_after_split(*this, split, top,
child_stack_entry[0],
child_stack_entry[1],
features,
labels);
// Update the Child entries of the parent
// Using InteriorNodeBase because exact parameter form not needed.
// look at the Node base before getting scared.
last_node_pos = topology_.size();
if(top.leftParent != StackEntry_t::DecisionTreeNoParent)
{
NodeBase(topology_,
parameters_,
top.leftParent).child(0) = last_node_pos;
}
else if(top.rightParent != StackEntry_t::DecisionTreeNoParent)
{
NodeBase(topology_,
parameters_,
top.rightParent).child(1) = last_node_pos;
}
// Supply the split functor with the Node type it requires.
// set the address to which the children of this node should point
// to and push back children onto stack
if(!isLeafNode(NodeID))
{
child_stack_entry[0].leftParent = topology_.size();
child_stack_entry[1].rightParent = topology_.size();
child_stack_entry[0].rightParent = -1;
child_stack_entry[1].leftParent = -1;
stack.push_back(child_stack_entry[0]);
stack.push_back(child_stack_entry[1]);
}
//copy the newly created node form the split functor to the
//decision tree.
NodeBase node(split.createNode(), topology_, parameters_ );
}
if(garbaged_child!=-1)
{
Node<e_ConstProbNode>(topology_,parameters_,garbaged_child).copy(Node<e_ConstProbNode>(topology_,parameters_,last_node_pos));
int last_parameter_size = Node<e_ConstProbNode>(topology_,parameters_,garbaged_child).parameters_size();
topology_.resize(last_node_pos);
parameters_.resize(parameters_.size() - last_parameter_size);
if(top.leftParent != StackEntry_t::DecisionTreeNoParent)
NodeBase(topology_,
parameters_,
top.leftParent).child(0) = garbaged_child;
else if(top.rightParent != StackEntry_t::DecisionTreeNoParent)
NodeBase(topology_,
parameters_,
top.rightParent).child(1) = garbaged_child;
}
}
} //namespace detail
} //namespace vigra
#endif //VIGRA_RANDOM_FOREST_DT_HXX
|