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
|
#include <core/optimizer.hpp>
#include <methods/powell/powell_method.hpp>
#include <methods/aga/aga.hpp>
#include <error_estimator/error_estimator.hpp>
#include <core/fitter.hpp>
#include <vector>
#include <iostream>
#include <data_sets/default_data_set.hpp>
#include <statistics/chisq.hpp>
using namespace opt_utilities;
//declear a class derived from func_obj
class lin1d
:public opt_utilities::model<double,double,std::vector<double>,std::string>
{
private:
model<double,double,std::vector<double> >* do_clone()const
{
return new lin1d(*this);
}
public:
lin1d()
{
this->push_param_info(param_info<std::vector<double> >("k",1,-10,10));
this->push_param_info(param_info<std::vector<double> >("b",0,-10,10));
}
public:
double do_eval(const double& x,const std::vector<double>& param)
{
return x*get_element(param,0)+get_element(param,1);
}
private:
std::string do_to_string()const
{
return "linear model\n"
"y=k*x+b\n";
}
};
int main()
{
fitter<double,double,std::vector<double>,double,std::string> f;
f.set_model(lin1d());
f.set_opt_method(powell_method<double,std::vector<double> >());
//f.set_opt_method(aga_method<double,std::vector<double> >());
default_data_set<double,double> ds;
for(int i=0;i<100;++i)
{
ds.add_data(data<double,double>(i,i*3+5,.1,.1,0,0));
}
f.load_data(ds);
f.set_statistic(chisq<double,double,std::vector<double>,double,std::string>());
f.fit();
cout<<f.get_param_value("k")<<endl;
cout<<f.get_param_value("b")<<endl;
double lower=3;
double upper=6;
estimate_error(f,std::string("k"),lower,upper,1.,1E-10);
std::cout<<lower<<"\t"<<upper<<endl;
}
|
