linear core More...

#include <core.h>

Inheritance diagram for linearCore:

Public Member Functions
	linearCore (const PsiData *data=NULL, const int sigmoid=1, const double alpha=0.1)
	construcor
	linearCore (const linearCore &original)
	copy constructor
double	g (double x, const std::vector< double > &prm) const
	evaluate the core of the sigmoid
double	dg (double x, const std::vector< double > &prm, int i) const
	first derivative w.r.t. parameter i
double	dgx (double x, const std::vector< double > &prm) const
	evaluate the first derivative of the core with respect to stimulus intensity
double	ddg (double x, const std::vector< double > &prm, int i, int j) const
	second derivative w.r.t. parameters i and j
double	inv (double y, const std::vector< double > &prm) const
	inverse of the core
double	dinv (double y, const std::vector< double > &prm, int i) const
	deriviative of the inverse w.r.t. parameter i
std::vector< double >	transform (int nprm, double a, double b) const
	transform logistic regression parameters to useful ones for this core
PsiCore *	clone (void) const
	clone object by value
Static Public Member Functions
static std::string	getDescriptor (void)
	get a short string that identifies the type of core

Detailed Description

linear core

The core of the sigmoid is simply a*x+b, where a and b are the first two parameters. This is the parameterization that would be used in the context of generalized linear models. The parameters do not have an obvious interpretation in terms of psychophysically meaningful quantities. However, it might well be that in this form, the parameters are more independent, which is particularly important for MCMC.

Constructor & Destructor Documentation

linearCore::linearCore	(	const PsiData *	data = `NULL`,
		const int	sigmoid = `1`,
		const double	alpha = `0.1`
	)			`[inline]`

construcor

Parameters:

	data	ignored
	sigmoid	ignored
	alpha	ignored

Member Function Documentation

double linearCore::ddg	(	double	x,
		const std::vector< double > &	prm,
		int	i,
		int	j
	)			const `[inline, virtual]`

second derivative w.r.t. parameters i and j

Parameters:

	x	stimulus intensity
	prm	parameter vector
	i	index of the parameter we want for the first derivative
	j	index of the parameter we want for the second derivative

Reimplemented from PsiCore.

double linearCore::dg	(	double	x,
		const std::vector< double > &	prm,
		int	i
	)			const `[inline, virtual]`

first derivative w.r.t. parameter i

Parameters:

	x	stimululs intensity
	prm	parameter vector
	i	index of the parameter we want the derivative to

Reimplemented from PsiCore.

double linearCore::dgx	(	double	x,
		const std::vector< double > &	prm
	)			const `[inline, virtual]`

evaluate the first derivative of the core with respect to stimulus intensity

Parameters:

	x	stimulus intensity
	prm	parameter vector

Reimplemented from PsiCore.

double linearCore::dinv	(	double	y,
		const std::vector< double > &	prm,
		int	i
	)			const `[inline, virtual]`

deriviative of the inverse w.r.t. parameter i

Parameters:

	y	value at which the derivative of the inverse should be evaluated
	prm	parameter vector
	i	index of the parameter we want the derivative to

Reimplemented from PsiCore.

double linearCore::g	(	double	x,
		const std::vector< double > &	prm
	)			const `[inline, virtual]`

evaluate the core of the sigmoid

Parameters:

	x	stimulus intensity
	prm	parameter vector

Reimplemented from PsiCore.

double linearCore::inv	(	double	y,
		const std::vector< double > &	prm
	)			const `[inline, virtual]`

inverse of the core

Parameters:

	y	value to be inverted
	prm	parameter vector

Reimplemented from PsiCore.

std::vector<double> linearCore::transform	(	int	nprm,
		double	a,
		double	b
	)			const `[inline, virtual]`

transform logistic regression parameters to useful ones for this core

Parameters:

	nprm	number of parameters in the whole model
	a	intercept parameter of the logistic regression model
	b	slope parameter of the logistic regression

Reimplemented from PsiCore.

The documentation for this class was generated from the following file:

src/core.h

linearCore Class Reference

Public Member Functions

Static Public Member Functions

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation