a struct for a signal sample for the pure CsI calorimeter More...

#include <EclConfigurationPure.h>

Public Member Functions
void	InitSample (const TH1F , const TH1F )
	initialisation of signal sample

double	Accumulate (const double a, const double t0, double *s) const

Public Attributes
double	m_sumscale
	energy deposit in fitting window scale factor

double	m_ft [m_nlPure *m_ns]
	Simulated signal shape.

double	m_ft1 [m_nlPure *m_ns]
	Simulated signal shape.

Detailed Description

a struct for a signal sample for the pure CsI calorimeter

Definition at line 55 of file EclConfigurationPure.h.

Member Function Documentation

◆ Accumulate()

double Accumulate	(	const double	a,
		const double	t0,
		double *	s
	)		const

Parameters

[in]	a	Signal amplitude
[in]	t0	Signal offset
[out]	s	Output array with added signal

Returns: Energy deposition in ADC units

Definition at line 57 of file EclConfigurationPure.cc.

{
  // input parameters
  // a -- signal amplitude
  // t -- signal offset
  // output parameter
  // s -- output array with added signal
  const double itick = 1 / getTickPure();          // reciprocal to avoid division in usec^-1 (has to be evaluated at compile time)
  const double  tlen = m_nlPure - 1.0 / m_ns;   // length of the sampled signal in ADC clocks units
  const double  tmax = m_tmin + m_nsmp - 1; // upper range of the fit region
 
  double t = t0 * itick; // rescale time in usec to ADC clocks
  double x0 = t, x1 = t + tlen;
 
  if (x0 > tmax) return 0; // signal starts after the upper range of output array -- do nothing
  if (x0 < m_tmin) {
    if (x1 < m_tmin) return 0; // signal ends before lower range of output array -- do nothing
    x0 = m_tmin; // adjust signal with range of output array
  }
 
  int imax = m_nsmp; // length of sampled signal is long enough so
  // the last touched element is the last element
  // of the output array
  if (x1 < tmax) { // if initial time is too early we need to adjust
    // the last touched element of output array to avoid
    // out-of-bound situation in m_ft
    imax = x1 - m_tmin; // imax is always positive so floor can be
    // replace by simple typecast
    imax += 1; // like s.end()
  }
 
  double epsilon = 1.0 / m_ns / 10.;
  double imind = ceil(x0 - m_tmin + epsilon); // store result in double to avoid int->double conversion below
  // the ceil function today at modern CPUs is surprisingly fast (before it was horribly slow)
  int imin = imind; // starting point to fill output array
  double w = ((m_tmin - t) + imind - 1) * double(m_ns);
  int jmin = w ; // starting point in sampled signal array
  w -= jmin;
 
  // use linear interpolation between samples. Since signal samples
  // are aligned with output samples only two weights are need to
  // calculate to fill output array
  const double w1 = a * w, w0 = a - w1;
  double sum = 0;
  //cout <<"Filling energy: " << a << " time " << t << endl;
  //cout <<"imin: " << imin << " imax: " << imax << endl;
  for (int i = imin, j = jmin; i < imax; i++, j += m_ns) {
    double amp = 0;
    if (j >= 0)  amp = w0 * m_ft[j] + w1 * m_ft[j + 1];
    //double amp = a * m_ft[j];
    // cout << i << ":" << j << " " << m_ft[j] << " " << w * m_ft[j] + (1-w) * m_ft[j+1] << endl;
    s[i] += amp;
    sum  += amp;
  }
  //cout << endl;
  return sum * m_sumscale;
}

◆ InitSample()

void InitSample	(	const TH1F *	sampledfun,
		const TH1F *	sampledfunDerivative
	)

initialisation of signal sample

Definition at line 26 of file EclConfigurationPure.cc.

{
  const int N = m_ns * m_nlPure;
  double r1 = 32 / m_ns;
  double r2 = EclConfiguration::getTick() / EclConfiguration::m_ntrg * 8 / getTickPure();
 
  ECLSampledShaper dsp(sampledfun, round(r1 / r2));
  dsp.fillarray(N, m_ft);
  double maxval = * max_element(m_ft, m_ft + N);
  double scale = 1.0 / maxval;
  for (int i = 0; i < N; ++i)
    m_ft1[i] = m_ft1[i] * scale;
  double maxval2 = * max_element(m_ft, m_ft + N);
  assert(maxval2 - 1.0 < 0.001);
  double sum = 0;
  for (int i = 0; i < N; i++) sum += m_ft[i];
  m_sumscale = m_ns / sum;
  //  for (int i = 0; i < N; ++i) m_ft1[i] = sampledfunDerivative->GetBinContent(i + 1);
  ECLSampledShaper dsp1(sampledfunDerivative, round(r1 / r2));
  dsp1.fillarray(N, m_ft1);
  scale = (r1 / r2) / maxval;
  for (int i = 0; i < N; ++i)
    m_ft1[i] = m_ft1[i] * scale;
}

Member Data Documentation

◆ m_ft

double m_ft[m_nlPure *m_ns]

Simulated signal shape.

Definition at line 57 of file EclConfigurationPure.h.

◆ m_ft1

double m_ft1[m_nlPure *m_ns]

Simulated signal shape.

Definition at line 58 of file EclConfigurationPure.h.

◆ m_sumscale

double m_sumscale

energy deposit in fitting window scale factor

Definition at line 56 of file EclConfigurationPure.h.

The documentation for this struct was generated from the following files:

ecl/digitization/include/EclConfigurationPure.h
ecl/digitization/src/EclConfigurationPure.cc

Public Member Functions

Public Attributes