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
|
#ifndef BFGS_METHOD
#define BFGS_METHOD
#define OPT_HEADER
#include <core/optimizer.hpp>
//#include <blitz/array.h>
#include <limits>
#include <cstdlib>
#include <core/opt_traits.hpp>
#include <math/num_diff.hpp>
#include <cassert>
#include <cmath>
#include <ctime>
#include <vector>
#include <algorithm>
/*
*
*/
#include <iostream>
using std::cout;
using std::endl;
namespace opt_utilities
{
template <typename rT,typename pT>
class bfgs_method
:public opt_method<rT,pT>
{
private:
pT start_point;
rT threshold;
func_obj<rT,pT>* p_fo;
optimizer<rT,pT>* p_optimizer;
typedef typename element_type_trait<pT>::element_type element_type;
element_type* mem_pool;
element_type** invBk;
element_type** invBk1;
bool bstop;
private:
const char* do_get_type_name()const
{
return "asexual genetic algorithm";
}
rT func(const pT& x)
{
assert(p_fo!=0);
return p_fo->eval(x);
}
public:
bfgs_method()
:p_fo(0),p_optimizer(0),
invBk(0),invBk1(0);
{
}
virtual ~bfgs_method()
{
destroy_workspace();
};
bfgs_method(const bfgs_method<rT,pT>& rhs)
:p_fo(rhs.p_fo),p_optimizer(rhs.p_optimizer),
threshold(rhs.threshold)
{
}
bfgs_method<rT,pT>& operator=(const bfgs_method<rT,pT>& rhs)
{
p_fo=rhs.p_fo;
p_optimizer=rhs.p_optimizer;
threshold=rhs.threshold;
}
opt_method<rT,pT>* do_clone()const
{
return new bfgs_method<rT,pT>(*this);
}
void init_workspace(int n)
{
destroy_workspace();
mem_pool=new element_type[n*n*2];
invBk=new element_type*[n];
invBk1=new element_type*[n];
for(size_t i=0;i!=n;++i)
{
invBk[i]=mem_pool+i*n;
invBk1[i]=invBk[i]+n*n;
}
for(size_t i=0;i!=n;++i)
{
for(size_t j=0;j!=n;++j)
{
invBk[i][j]=(i==j?1:0);
}
}
}
void destroy_workspace()
{
delete[] mem_pool;
delete[] invBk;
delete[] invBk1;
}
public:
void do_set_start_point(const pT& p)
{
start_point=p;
init_workspace(get_size(p));
}
pT do_get_start_point()const
{
}
void do_set_precision(rT t)
{
threshold=t;
}
rT do_get_precision()const
{
return threshold;
}
void do_set_optimizer(optimizer<rT,pT>& o)
{
p_optimizer=&o;
p_fo=p_optimizer->ptr_func_obj();
}
pT do_optimize()
{
pT s;
resize(s,get_size(start_point));
pT old_grad;
pT y;
resize(old_grad,get_size(start_point));
resize(y,get_size(start_point));
for(size_t i=0;i!=get_size(p);++i)
{
set_element(old_grad,i,gradient(*p_fo,start_point,i));
set_element(s,i,0);
for(size_t j=0;j!=get_size(p);++j)
{
s[i]+=invBk[i][j]*old_grad[j];
}
}
linmin(start_point,s,*p_fo);
for(size_t i=0;i!=get_size(p);++i)
{
set_element(y,gradient(*p_fo,start_point,i)-get_element(old_grad,i));
}
rT sy=0;
pT invBy;
pT yinvB;
resize(invBy,get_size(p));
resize(yinvB,get_size(p));
for(size_t i=0;i!=get_size(p);++i)
{
sy+=s[i]*y[i];
for(size_t j=0;j!=get_size(p);++j)
{
invBy[i]+=invBk[i][j]*y[j];
yinvB[i]+=y[j]*invBk[j][i];
}
}
rT yinvBy=0;
for(size_t i=0;i!=get_size(p);++i)
{
yinvBy+=invBy[i]*y[i];
}
for(size_t i=0;i<get_size(p);++i)
{
for(size_t j=0;j<get_size(p);++j)
{
invBk[i][j]+=((sy+yinvBy)*s[i]*s[j]/(sy*sy)-(invBy[i]*s[j]+s[i]*yinvB[j])/(sy));
}
}
}
void do_stop()
{
bstop=true;
}
};
}
#endif
//EOF
|
