aboutsummaryrefslogtreecommitdiffstats
path: root/pre_estimaters/vgauss1d_estimater.hpp
blob: e3d1526f0acaa1472f47d0eaec785c72671263cf (plain)
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
#ifndef VGAUSS1D_ESTIMATER
#define VGAUSS1D_ESTIMATER
#include <core/pre_estimater.hpp>
#include <misc/optvec.hpp>
#include <vmodels/gauss1d.hpp>
#include <vector>

namespace opt_utilities
{
  template <typename T>
  class gauss1d_estimater
    :public pre_estimater<optvec<T>,optvec<T>,optvec<T>,std::string>
  {
  public:
    gauss1d_estimater()
    {
      this->set_model_id("1d gaussian");
    }

    gauss1d_estimater<T>* do_clone()const
    {
      return new gauss1d_estimater<T>(*this);
    }

    void do_estimate(const data_set<optvec<T>,optvec<T> >& d,model<optvec<T>,optvec<T>,optvec<T>,std::string>& m)const
    {
      int n=d.size();
      T xmean=0;
      T x2mean=0;
      T wgt=0;
      T wgt2=0;
      for(int i=0;i<n;++i)
	{
	  T x=d.get_data(i).get_x()[0];
	  T y=d.get_data(i).get_y()[0];
	  xmean+=x*y;
	  x2mean+=x*x*y;
	  wgt+=y;
	}
      xmean/=wgt;
      x2mean/=wgt;
      T sigma=std::sqrt(x2mean-xmean*xmean);
      m.set_param_value("x0",xmean);
      m.set_param_value("sigma",sigma);
      
    }
  };
}



#endif