Derived class of class Func representing a Mixture of several distinct functions. The function is of the form ln( g(r) / h(r) ), where g(r) is a mixture of a Gaussian and Exponential distributions, and h(r) is a Gaussian distribution. See methods and implementation for more information. More...

#include <MixtureFunc.hh>

Inheritance diagram for core::scoring::constraints::MixtureFunc:

[legend]

Collaboration diagram for core::scoring::constraints::MixtureFunc:

[legend]

List of all members.

Public Member Functions
	MixtureFunc (Real const anchor, Real const gaussian_param, Real const exp_param, Real const mixture_param, Real const bg_mean, Real const bg_sd)
FuncOP	clone () const
	returns a clone of this MixtureFunc
Real	func (Real const x) const
	Returns the value of this MixtureFunc evaluated at distance x.
Real	dfunc (Real const x) const
	Returns the value of the first derivative of this MixtureFunc at distance x.
virtual void	show_definition (std::ostream &out) const
	show the definitio of this MixtureFunc to the specified output stream.
Real	dfunc_component (Real const x, Real &g, Real &h, Real &g_prime, Real &h_prime) const
	Function that's used for debugging. Given x, this calculates g(x), h(x), g'(x) and h'(x).
Real	calc_kl_divergence () const
	Calculates the K-L divergence between the inferred and background distributions.
virtual void	show (std::ostream &out) const
	Prints this MixtureFunc to the given ostream.
void	read_data (std::istream &in)
	The parameters are:
Real	func_ (Real x) const
	Returns the value of this MixtureFunc evaluated at distance x.
Friends
std::ostream &	operator<< (std::ostream &out, const MixtureFunc &f)
	Calls show( out ) on this MixtureFunc.

Detailed Description

Derived class of class Func representing a Mixture of several distinct functions. The function is of the form ln( g(r) / h(r) ), where g(r) is a mixture of a Gaussian and Exponential distributions, and h(r) is a Gaussian distribution. See methods and implementation for more information.

Constructor & Destructor Documentation

core::scoring::constraints::MixtureFunc::MixtureFunc	(	Real const	anchor,
		Real const	gaussian_param,
		Real const	exp_param,
		Real const	mixture_param,
		Real const	bg_mean,
		Real const	bg_sd
	)		`[inline]`

Constructor for MixtureFunc. Arguments to the constructor are:

anchor: parameter representing the value at which this function is anchored, represents the mean of the Gaussian distribution and the highest point of the exponential distribution.
gaussian_param: parameter for Gaussian portion of g(r), representing the standard deviation of a Gaussian distribution around anchor.
exp_param: parameter for Exponential portion of g(r), representing the rate at which the exponential distribution drops off from anchor.
mixture_param: parameter describing the mixture of the Gaussian and Exponential functions that make up g(r).
bg_mean: parameter representing the mean of h(r).
bg_sd: parameter representing the standard deviation of h(r).

Referenced by clone().

Member Function Documentation

Real core::scoring::constraints::MixtureFunc::calc_kl_divergence ( ) const

Calculates the K-L divergence between the inferred and background distributions.

References core::sequence::end.

FuncOP core::scoring::constraints::MixtureFunc::clone ( ) const [inline, virtual]

returns a clone of this MixtureFunc

Implements core::scoring::constraints::Func.

References MixtureFunc().

Real core::scoring::constraints::MixtureFunc::dfunc ( Real const x ) const [virtual]

Returns the value of the first derivative of this MixtureFunc at distance x.

Implements core::scoring::constraints::Func.

Real core::scoring::constraints::MixtureFunc::dfunc_component	(	Real const	x,
		Real &	g,
		Real &	h,
		Real &	g_prime,
		Real &	h_prime
	)		const

Function that's used for debugging. Given x, this calculates g(x), h(x), g'(x) and h'(x).

References core::scoring::constraints::dexponential(), core::scoring::constraints::dgaussian(), core::scoring::constraints::exponential_deriv(), and core::scoring::constraints::gaussian_deriv().

Real core::scoring::constraints::MixtureFunc::func ( Real const x ) const [virtual]

Returns the value of this MixtureFunc evaluated at distance x.

Implements core::scoring::constraints::Func.

Real core::scoring::constraints::MixtureFunc::func_ ( Real x ) const

Returns the value of this MixtureFunc evaluated at distance x.

private

References core::scoring::constraints::dexponential(), and core::scoring::constraints::dgaussian().

void core::scoring::constraints::MixtureFunc::read_data ( std::istream & in ) [virtual]

The parameters are:

Initializes this MixtureFunc from the given istream. An example of the type of string from which the istream should be constructed is: "MIXTUREFUNC 6.9734 3.598 0.222 0.872 19.396 7.643". The interpretation is to create initialize this MixtureFunc object with the following parameters: