mirror of
https://github.com/slsdetectorgroup/slsDetectorPackage.git
synced 2025-04-21 19:30:03 +02:00
157 lines
3.6 KiB
C++
Executable File
157 lines
3.6 KiB
C++
Executable File
#ifndef MOVINGSTAT_H
|
|
#define MOVINGSTAT_H
|
|
|
|
#include <math.h>
|
|
|
|
|
|
class MovingStat
|
|
{
|
|
|
|
/** @short approximated moving average structure */
|
|
public:
|
|
|
|
|
|
/** constructor
|
|
\param nn number of samples parameter to be used
|
|
*/
|
|
MovingStat(int nn=1000) : n(nn), m_n(0) {}
|
|
|
|
/**
|
|
clears the moving average number of samples parameter, mean and standard deviation
|
|
*/
|
|
void Clear()
|
|
{
|
|
m_n = 0;
|
|
m_newM=0;
|
|
m_newM2=0;
|
|
}
|
|
|
|
/**
|
|
clears the moving average number of samples parameter, mean and standard deviation
|
|
*/
|
|
void Set(double val, double rms=0, int m=-1)
|
|
{
|
|
if (m>=0) m_n = m; else m_n = n;
|
|
m_newM=val*m_n;
|
|
SetRMS(rms);
|
|
}
|
|
/**
|
|
clears the moving average number of samples parameter, mean and standard deviation
|
|
*/
|
|
void SetRMS(double rms)
|
|
{
|
|
if (rms<=0) {
|
|
m_newM2=m_newM*m_newM/n;
|
|
m_n=0;
|
|
} else {
|
|
if (m_n>0)
|
|
m_newM2=(m_n*rms*rms+m_newM*m_newM/m_n);
|
|
else {
|
|
m_newM2=(m_n*rms*rms+m_newM*m_newM/n);
|
|
m_n=0;
|
|
}
|
|
}
|
|
}
|
|
|
|
/** sets number of samples parameter
|
|
\param i number of samples parameter to be set
|
|
*/
|
|
|
|
int SetN(int i) {if (i>=1) n=i; return n;};
|
|
|
|
/**
|
|
gets number of samples parameter
|
|
\returns actual number of samples parameter
|
|
*/
|
|
int GetN() {return m_n;};
|
|
|
|
/** calculates the moving average i.e. adds if number of elements is lower than number of samples parameter, pushes otherwise
|
|
\param x value to calculate the moving average
|
|
*/
|
|
inline void Calc(double x) {
|
|
if (m_n<n) Add(x);
|
|
else Push(x);
|
|
}
|
|
/** adds the element to the accumulated average and standard deviation
|
|
\param x value to add
|
|
*/
|
|
inline void Add(double x) {
|
|
m_n++;
|
|
|
|
if (m_n == 1)
|
|
{
|
|
m_newM = x;
|
|
m_newM2 = x*x;
|
|
} else {
|
|
m_newM = m_newM + x;
|
|
m_newM2 = m_newM2 + x*x;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
inline void Push(double x)
|
|
{
|
|
/** adds the element to the accumulated average and squared mean, while subtracting the current value of the average and squared average
|
|
\param x value to push
|
|
*/
|
|
if (m_n == 0)
|
|
{
|
|
m_newM = x;
|
|
m_newM2 = x*x;
|
|
m_n++;
|
|
} else {
|
|
m_newM = m_newM + x - m_newM/m_n;
|
|
m_newM2 = m_newM2 + x*x - m_newM2/m_n;
|
|
}
|
|
|
|
}
|
|
|
|
/** returns the current number of elements of the moving average
|
|
\returns returns the current number of elements of the moving average
|
|
*/
|
|
int NumDataValues() const
|
|
{
|
|
return m_n;
|
|
}
|
|
/** returns the mean, 0 if no elements are inside
|
|
\returns returns the mean
|
|
*/
|
|
inline double Mean() const
|
|
{
|
|
return (m_n > 0) ? m_newM/m_n : 0.0;
|
|
}
|
|
|
|
/** returns the squared mean, 0 if no elements are inside
|
|
\returns returns the squared average
|
|
*/
|
|
double M2() const
|
|
{
|
|
return ( (m_n > 1) ? m_newM2/m_n : 0.0 );
|
|
}
|
|
|
|
/** returns the variance, 0 if no elements are inside
|
|
\returns returns the variance
|
|
*/
|
|
inline double Variance() const
|
|
{
|
|
return ( (m_n > 1) ? (M2()-Mean()*Mean()) : 0.0 );
|
|
}
|
|
|
|
/** returns the standard deviation, 0 if no elements are inside
|
|
\returns returns the standard deviation
|
|
*/
|
|
inline double StandardDeviation() const
|
|
{
|
|
return ( (Variance() > 0) ? sqrt( Variance() ) : -1 );
|
|
}
|
|
|
|
private:
|
|
int n; /**< number of samples parameter */
|
|
int m_n; /**< current number of elements */
|
|
double m_newM; /**< accumulated average */
|
|
double m_newM2; /**< accumulated squared average */
|
|
};
|
|
#endif
|