Tortik/resfitter.h at master · mefisto545/Tortik · GitHub

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
#ifndef RESFITTER_H
#define RESFITTER_H

#define DEF_MAX_STEPS 1e2;
#define DEF_MIN_ERROR 1e-3;
#define DEF_MIN_NORM_DIFF 1e-3;
#define NOISE_LEVEL 0.25;

#include <vector>
#include "filedata.h"
#include <math.h>
#include <stack>
#include <iostream>
#include <QTime>

#include "Libs/Eigen/unsupported/Eigen/Polynomials"
#include <algorithm>
#include <cmath>
#include <complex>
#include <iterator>

using namespace std;

class ResFitter
{
    int maxSteps;
    double minError;
    double step;
    double minNormDiff;
    vector<double> errors;
    vector<double> freq;
    vector<double> theta;
    vector<struct Resonance> resData;
    Resonance resonance;
    double params[4] = {-1.0,-1.0,-1.0,-1.0};
    FileData *file;

    double lorentzDyc(double x, double xc, double w);
    double lorentzDyc(double x, double *params);

    double lorentzDxc(double x, double y0, double yc, double xc, double w);
    double lorentzDxc(double x, double *params);

    double lorentzDw(double x, double y0, double yc, double xc, double w);
    double lorentzDw(double x, double *params);

    double errorMSE(double *params);
    void gradDescentStep(double *params);
    void stochGradDescentStep(double *params, int numPoint);
    void gradDescent();
    bool readDataFromStack(stack <struct Resonance> & stack);
    void findInitialParams(double y0);
    double findStep(double* initialParams, double* gradient, int numPoint);

    vector<double> eigenSolver(const std::vector<double>& coeffA);
    vector<double> eigenSolverMonic(const std::vector<double>& coefficients);
    int randInt();
public:
    vector<struct Resonance> fittedData;
    /*ResFitter - method that perform fitting data with Lorentz peak function. The result of fitting writes in text files.
     * maxSteps - maximum number of steps that can be performed in gradient descend
     * minError - the minimum error level that will stop the descend
     * step - the value in gradient descend method  y1 = y0 - step*grad
     * file - element of class, that contain the fitting data*/
    ResFitter(double maxSteps, double minError, FileData *file);
    static double lorentz(double x, double *params);
    void fitData(stack <struct Resonance> &stack, const vector<double> &baseline);
    static double lorentz(double x, double y0, double yc, double xc, double w);
};

#endif // RESFITTER_H