hitfit::Resolution Class Reference

#include <Resolution.h>

Public Member Functions
double	C () const
	Return the C term (constant term) More...
bool	inverse () const
	Return the setting of the inverse flag. More...
double	m () const
	Return the exponent factor in the resolution term. More...
double	N () const
	Return the N term (noise term) More...
double	pick (double x, double p, CLHEP::HepRandomEngine &engine) const
	Generate random value from a Gaussian distribution described by this resolution. Given a value , measured for an object with momentum , pick a new value from a Gaussian distribution described by this resolution: with mean and width . More...
double	R () const
	Return the R term (resolution term) More...
	Resolution (double C, double R, double m, double N, bool inverse=false)
	Constructor to initialize with four values for C, R, m, N, and the boolean for inverse. More...
	Resolution (double res, bool inverse=false)
	Constructor to initialize a constant resolution. More...
	Resolution (std::string s="")
	Constructor, initialize from a string. More...
double	sigma (double p) const
	Return the uncertainty for a variable with magnitude p. More...

Private Attributes
double	_constant_sigma
bool	_inverse
double	_noise_sigma
double	_resolution_exponent
double	_resolution_sigma

Friends
std::ostream &	operator<< (std::ostream &s, const Resolution &r)
	Output stream operator, print the content of this Resolution to an output stream. More...

Detailed Description

@class Resolution

@brief Calculate and represent resolution for a physical quantity.

This class calculate resolutions for some quantity.  In general we have
three parameters:
- <b>C</b> constant term.
- <b>R</b> resolution term.
- <b>N</b> noise term.

Given a physical quantitiy \f$p\f$, we calculate the uncertainty
in quantity \f$x\f$ as:

\f[
\sigma(x) = \sqrt{C^{2}p^{2} + R^{2}p + N^{2}}
\f]

In addition, we also have an inverse flag.  If the flag is set, we
take the inverse of \f$p\f$ before doing the calculations.
Therefore \f$\sigma(x)\f$ is regarded as actually \f$\sigma(1/x)\f$.

We encode he resolution parameters into a string, from which these
objects get initialized.  The format is

\verbatim

[-]C[,R[,N]]

where parameters within the brackets are optional. If the leading minus is present, the inverse flag is turned on. Omitted parameters are set to 0.

Definition at line 98 of file Resolution.h.

Constructor & Destructor Documentation

Resolution() [1/3]

hitfit::Resolution::Resolution

(

std::string

s = ""

)

Constructor, initialize from a string.

Parameters

s	A string encoding the resolution parameters, as described in the class description.

Definition at line 82 of file Resolution.cc.

      : _resolution_exponent(0) {
    _inverse = false;
    _constant_sigma = 0;
    _resolution_sigma = 0;
    _noise_sigma = 0;
 
    // Skip spaces.
    double x;
    string::size_type i = 0;
    while (i < s.size() && isspace(s[i]))
      ++i;
 
    // Check for the inverse flag.
    if (s[i] == '-') {
      _inverse = true;
      ++i;
    } else if (s[i] == '+') {
      ++i;
    }
 
    // Get the constant term.
    if (get_field(s, i, x))
      _constant_sigma = x;
    i = s.find(',', i);
 
    // Look for a resolution term.
    if (i != string::npos) {
      ++i;
      if (get_field(s, i, x))
        _resolution_sigma = x;
 
      // Look for a noise term.
      i = s.find(',', i);
      if (i != string::npos) {
        if (get_field(s, i + 1, x))
          _noise_sigma = x;
      }
    }
  }

References _constant_sigma, _inverse, _noise_sigma, _resolution_sigma, mps_fire::i, and alignCSCRings::s.

Resolution() [2/3]

hitfit::Resolution::Resolution	(	double	C,
		double	R,
		double	m,
		double	N,
		bool	inverse = false
	)

Constructor to initialize with four values for C, R, m, N, and the boolean for inverse.

Parameters

C	The constant term
R	The resolution term
m	The exponent factor term
N	The noise term
inverse	The inverse flag.

Definition at line 130 of file Resolution.cc.

      : _constant_sigma(C), _resolution_sigma(R), _resolution_exponent(m), _noise_sigma(N), _inverse(inverse) {}

Resolution() [3/3]

hitfit::Resolution::Resolution	(	double	res,
		bool	inverse = false
	)

Constructor to initialize a constant resolution.

Parameters

res	The resolution value.
inverse	The inverse flag.

Definition at line 133 of file Resolution.cc.

      : _constant_sigma(0), _resolution_sigma(0), _resolution_exponent(0), _noise_sigma(res), _inverse(inverse) {}

Member Function Documentation

C()

double hitfit::Resolution::C

(

)

const

Return the C term (constant term)

Definition at line 155 of file Resolution.cc.

{ return _constant_sigma; }

References _constant_sigma.

inverse()

bool hitfit::Resolution::inverse

(

)

const

Return the setting of the inverse flag.

Definition at line 144 of file Resolution.cc.

  {
    return _inverse;
  }

References _inverse.

m()

double hitfit::Resolution::m

(

)

const

Return the exponent factor in the resolution term.

Definition at line 159 of file Resolution.cc.

{ return _resolution_exponent; }

References _resolution_exponent.

N()

double hitfit::Resolution::N

(

)

const

Return the N term (noise term)

Definition at line 161 of file Resolution.cc.

{ return _noise_sigma; }

References _noise_sigma.

pick()

double hitfit::Resolution::pick	(	double	x,
		double	p,
		CLHEP::HepRandomEngine &	engine
	)		const

Generate random value from a Gaussian distribution described by this resolution. Given a value , measured for an object with momentum , pick a new value from a Gaussian distribution described by this resolution: with mean and width .

Parameters

x	The quantity value (distributed mean).
p	The momentum, for calculating the width.
engine	The underlying random number generator.

Definition at line 181 of file Resolution.cc.

  {
    CLHEP::RandGauss gen(engine);
    if (_inverse)
      return 1 / gen.fire(1 / x, sigma(p));
    else
      return gen.fire(x, sigma(p));
  }

References relval_steps::gen(), and AlCaHLTBitMon_ParallelJobs::p.

Referenced by hitfit::Lepjets_Event::smear().

R()

double hitfit::Resolution::R

(

)

const

Return the R term (resolution term)

Definition at line 157 of file Resolution.cc.

{ return _resolution_sigma; }

References _resolution_sigma.

sigma()

double hitfit::Resolution::sigma

(

double

p

)

const

Return the uncertainty for a variable with magnitude p.

Parameters

p	The momentum.

Definition at line 163 of file Resolution.cc.

  {
    if (_inverse)
      p = 1 / p;
 
    return sqrt((_constant_sigma * _constant_sigma * p + _resolution_sigma * _resolution_sigma) * p +
                _noise_sigma * _noise_sigma);
  }

References AlCaHLTBitMon_ParallelJobs::p, and mathSSE::sqrt().

Friends And Related Function Documentation

operator<<

std::ostream& operator<< ( std::ostream &  s, const Resolution &  r  )

friend

Output stream operator, print the content of this Resolution to an output stream.

Parameters

s	The stream to which to write.
r	The instance of Resolution to be printed.

Definition at line 211 of file Resolution.cc.

  {
    if (r._inverse)
      s << "-";
    s << r._constant_sigma << "," << r._resolution_sigma << "," << r._noise_sigma;
    return s;
  }

Member Data Documentation

_constant_sigma

double hitfit::Resolution::_constant_sigma

private

The constant term.

Definition at line 202 of file Resolution.h.

Referenced by C(), and Resolution().

_inverse

bool hitfit::Resolution::_inverse

private

The inverse flag.

Definition at line 222 of file Resolution.h.

Referenced by inverse(), and Resolution().

_noise_sigma

double hitfit::Resolution::_noise_sigma

private

The noise term.

Definition at line 217 of file Resolution.h.

Referenced by N(), and Resolution().

_resolution_exponent

double hitfit::Resolution::_resolution_exponent

private

The m exponential factor in the resolution term.

Definition at line 212 of file Resolution.h.

Referenced by m().

_resolution_sigma

double hitfit::Resolution::_resolution_sigma

private

The resolution term.

Definition at line 207 of file Resolution.h.

Referenced by R(), and Resolution().