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
|
/**
* Define decision forest as a vector of decision tree's pointer,
* testing processing runs in parallel by openMP.
*/
#ifndef FOREST_H
#define FOREST_H
#include "tree.h"
template<class S, class C, class dataT>
class DecisionForest
{
public:
typedef DecisionTree<S, C, dataT> DecisionTreeT;
DecisionForest(): verbose_(false) {}
DecisionForest(bool verbose): verbose_(verbose) {}
~DecisionForest()
{
for (size_t i = 0; i < trees_.size(); ++i)
{
delete trees_[i];
}
}
DecisionTreeT* AddTree()
{
DecisionTreeT* ctree = new DecisionTreeT(verbose_);
ctree->idx_ = trees_.size();
trees_.push_back(ctree);
return ctree;
}
void Apply(MLData<dataT, S*>& testingData,
Vector<Vector<S*> >& testingResult)
{
size_t treeNum = trees_.size();
size_t dataNum = testingData.Size();
testingResult.Resize(treeNum);
#pragma omp parallel for
for (index_t i = 0; i < treeNum; ++i)
{
testingResult[i].Resize(dataNum);
trees_[i]->Apply(testingData, testingResult[i]);
}
}
// Added by Paul. This version of the method allows the same testingResult
// vector to be reused. For speed, it does not check that the size of the
// testing result is sufficient
// This version of the method takes a preallocated vector
// index and a preallocated vector response_vec of the same size as
// testingData. It does not do allocation to save on computation time
void ApplyFast(MLData<dataT, S*>& testingData,
Vector<Vector<S*> >& testingResult,
std::vector<index_t> &index,
std::vector<bool> &response)
{
size_t treeNum = trees_.size();
for (index_t i = 0; i < treeNum; ++i)
trees_[i]->ApplyFast(testingData, testingResult[i], index, response);
}
void Print(int level)
{
for (index_t i = 0; i < trees_.size(); ++i)
{
std::cout << "tree " << i << " information:" << std::endl;
trees_[i]->Print(level);
}
}
void Read(std::istream& is)
{
size_t treeNum = 0;
readBasicType(is, verbose_);
readBasicType(is, treeNum);
size_t treeNumBefore = trees_.size();
if (treeNumBefore != 0)
{
for (index_t i = 0; i < treeNumBefore; ++i)
{
delete trees_[i];
}
}
trees_.resize(treeNum);
for (index_t i = 0; i < treeNum; ++i)
{
trees_[i] = new DecisionTreeT(verbose_);
trees_[i]->Read(is);
}
}
void Write(std::ostream& os)
{
size_t treeNum = trees_.size();
writeBasicType(os, verbose_);
writeBasicType(os, treeNum);
for (index_t i = 0; i < treeNum; ++i)
{
trees_[i]->Write(os);
}
}
int GetForestSize() { return trees_.size(); }
std::vector<DecisionTreeT*> trees_;
bool verbose_;
};
#endif // FOREST_H
|
