boundfun.cpp File Reference

(Last Changed on February 17, 2022 for git-revision e9b921f3a6f970f4a53398d141831583cf8746df.)

Code to implement placing constraints on parameters. See in Section 10.1 of the AUTODIF manual. The code in thse section of the AUTODIF manual appears to be out of date with respect to some of the code in this file. More...

#include "fvar.hpp"
#include <stdlib.h>
#include <stdio.h>
#include <math.h>

Go to the source code of this file.

Macros
#define	USE_BARD_PEN

Functions
dvariable	boundp (const prevariable &x, double fmin, double fmax, const prevariable &_fpen, double s)
	Compute penalty for exceeding bounds on parameter; variable ojbects. More...
dvariable	boundp (const prevariable &x, double fmin, double fmax, const prevariable &_fpen)
	Compute penalty for exceeding bounds on parameter; variable ojbects. More...
double	boundp (double x, double fmin, double fmax)
	Return scaled variable to original units; constant objects. More...
double	boundp (double x, double fmin, double fmax, const double &_fpen)
	Compute penalty for exceeding bounds on parameter; constant ojbects. More...
double	boundpin (double x, double fmin, double fmax, double s)
	Scale model variable over [-1,1]; constant objects. More...
double	boundpin (double x, double fmin, double fmax)
	Scale model variable over [-1,1]; constant objects. More...
double	boundpin (const prevariable &x, double fmin, double fmax, double s)
	Scale model variable over [-1,1]; variable objects. More...
double	boundpin (const prevariable &xx, double fmin, double fmax)
	Scale model variable over [-1,1]; variable objects. More...
dvariable	dfatan1 (dvariable x, double fmin, double fmax, const prevariable &_fpen)
	Scale input variable between upper and lower bounds and compute a penalty for exceeding the bounds. More...
dvariable	dfboundp (const prevariable &x, double fmin, double fmax)
	Computes the derivative of dvariable boundp(const prevariable& x, double fmin, double fmax,const prevariable& _fpen) More...
double	dftinv (double x, double fmin, double fmax)
	Inverse of dvariable dfatan1(dvariable x, double fmin, double fmax, const prevariable& _fpen) More...
double	dmax (double x, double y)
	Returns the maximum of two numbers. More...
double	dmin (double x, double y)
	Returns the minimum of two numbers. More...
double	nd2fboundp (double x, double fmin, double fmax, const double &fpen)
	Scale input variable between upper and lower bounds and compute a penalty for exceeding the bounds. More...
double	ndfboundp (double x, double fmin, double fmax, const double &fpen)
	Derivatative code for double nd2fboundp( double x, double fmin, double fmax,const double& fpen) More...

Detailed Description

Code to implement placing constraints on parameters. See in Section 10.1 of the AUTODIF manual. The code in thse section of the AUTODIF manual appears to be out of date with respect to some of the code in this file.

Author

David Fournier

Copyright

Copyright (c) 2008-2020 Regents of the University of California

Definition in file boundfun.cpp.

Macro Definition Documentation

#define USE_BARD_PEN

Use penalties recomended in Bard, Y. (1974) Nonlinear parameter estimation. Academic Press, New York.

Definition at line 18 of file boundfun.cpp.

Function Documentation

dvariable boundp	(	const prevariable &	x,
		double	fmin,
		double	fmax,
		const prevariable &	_fpen,
		double	s
	)

Compute penalty for exceeding bounds on parameter; variable ojbects.

Parameters

x	Variable scaled between [-1,1]
fmin	Lower bound of x
fmin	Upper bound of x
_fpen	On return, contains penalty if x > 1 or x < -1
s	Divide x by s before scaling and setting bounds

Returns

The variable x in original units

Definition at line 89 of file boundfun.cpp.

Referenced by boundp(), nd2fboundp(), funnel_init_df1b2variable::set_value(), funnel_init_bounded_df1b2vector::set_value(), random_effects_bounded_vector_info::set_value(), re_df1b2_init_bounded_vector::set_value(), set_value(), set_value_partial(), and param_init_bounded_matrix::shared_set_value().

dvariable boundp	(	const prevariable &	x,
		double	fmin,
		double	fmax,
		const prevariable &	_fpen
	)

Compute penalty for exceeding bounds on parameter; variable ojbects.

Parameters

x	Variable scaled between [-1,1]
fmin	Lower bound of x
fmin	Upper bound of x
_fpen	On return, contains penalty if x > 1 or x < -1

Returns

The variable x in original units

Definition at line 102 of file boundfun.cpp.

double boundp	(	double	x,
		double	fmin,
		double	fmax
	)

Return scaled variable to original units; constant objects.

Parameters

x	Variable scaled between [-1,1]
fmin	Lower bound of x
fmin	Upper bound of x

Returns

The variable x in original units

Definition at line 244 of file boundfun.cpp.

double boundp	(	double	x,
		double	fmin,
		double	fmax,
		const double &	_fpen
	)

Compute penalty for exceeding bounds on parameter; constant ojbects.

Inverse of double boundp(const double& x, double fmin, double fmax, const double& _fpen).

Parameters

x	Model variable
fmin	Lower bound of x
fmin	Upper bound of x
_fpen	On return, contains penalty if x > fmax or x < fmin

Returns

The variable x in original units

Definition at line 310 of file boundfun.cpp.

double boundpin	(	double	x,
		double	fmin,
		double	fmax,
		double	s
	)

Scale model variable over [-1,1]; constant objects.

Inverse of double boundp(double x, double fmin, double fmax)

Parameters

x	Model variable
fmin	Lower bound of x
fmin	Upper bound of x
s	Divide x by s before scaling and setting bounds

Returns

The variable x in original units

Definition at line 378 of file boundfun.cpp.

Referenced by boundpin(), param_init_bounded_vector::mc_scale(), param_init_bounded_number::mc_scale(), set_value_inv(), set_value_inv_partial(), param_init_bounded_matrix::shared_set_value_inv(), and funnel_init_df1b2variable::xinit().

double boundpin	(	double	x,
		double	fmin,
		double	fmax
	)

Scale model variable over [-1,1]; constant objects.

Inverse of double boundp(double x, double fmin, double fmax)

Parameters

x	Model variable
fmin	Lower bound of x
fmin	Upper bound of x

Returns

Scaled model variable over [-1,1].

Definition at line 391 of file boundfun.cpp.

double boundpin	(	const prevariable &	x,
		double	fmin,
		double	fmax,
		double	s
	)

Scale model variable over [-1,1]; variable objects.

Inverse of dvariable boundp(const prevariable& x, double fmin, double fmax,const prevariable& _fpen, double s).

Parameters

x	Model variable
fmin	Lower bound of x
fmin	Upper bound of x
s	Divide x by s before scaling and setting bounds

Returns

Scaled model variable over [-1,1].

Definition at line 434 of file boundfun.cpp.

double boundpin	(	const prevariable &	xx,
		double	fmin,
		double	fmax
	)

Scale model variable over [-1,1]; variable objects.

Inverse of dvariable boundp(const prevariable& x, double fmin, double fmax,const prevariable& _fpen).

Parameters

x	Model variable
fmin	Lower bound of x
fmin	Upper bound of x

Returns

Scaled model variable over [-1,1].

Definition at line 447 of file boundfun.cpp.

dvariable dfatan1	(	dvariable	x,
		double	fmin,
		double	fmax,
		const prevariable &	_fpen
	)

Scale input variable between upper and lower bounds and compute a penalty for exceeding the bounds.

Parameters

x	Variable to be scaled
fmin	Lower bound of x
fmin	Upper bound of x
_fpen	On return, contains penalty if x > fmax or x < fmin

Returns

Scaled value of x between fmin and fmax in the range [-1,1]

Definition at line 37 of file boundfun.cpp.

dvariable dfboundp	(	const prevariable &	x,
		double	fmin,
		double	fmax
	)

Computes the derivative of dvariable boundp(const prevariable& x, double fmin, double fmax,const prevariable& _fpen)

Parameters

x	Variable to be scaled
fmin	Lower bound of x
fmin	Upper bound of x

Returns

Derivative of scaled variable

Definition at line 170 of file boundfun.cpp.

Referenced by param_init_bounded_vector::sd_vscale(), param_init_bounded_number::sd_vscale(), and param_init_bounded_matrix::sd_vscale().

double dftinv	(	double	x,
		double	fmin,
		double	fmax
	)

Inverse of dvariable dfatan1(dvariable x, double fmin, double fmax, const prevariable& _fpen)

Parameters

x	Variable to be scaled
fmin	Lower bound of x
fmin	Upper bound of x

Definition at line 66 of file boundfun.cpp.

double dmax	(	double	x,
		double	y
	)

Returns the maximum of two numbers.

Parameters

x	First number
y	Second number

Definition at line 505 of file boundfun.cpp.

double dmin	(	double	x,
		double	y
	)

Returns the minimum of two numbers.

Parameters

x	First number
y	Second number

Definition at line 489 of file boundfun.cpp.

double nd2fboundp	(	double	x,
		double	fmin,
		double	fmax,
		const double &	fpen
	)

Scale input variable between upper and lower bounds and compute a penalty for exceeding the bounds.

Intended for use with df1b2 variables.

Parameters

x	Variable to be scaled
fmin	Lower bound of x
fmin	Upper bound of x
_fpen	On return, contains penalty if x > fmax or x < fmin

Returns

Scaled value of x between fmin and fmax in the range [-1,1]

Definition at line 282 of file boundfun.cpp.

Referenced by param_init_bounded_vector::curv_scale(), param_init_bounded_number::curv_scale(), and param_init_bounded_matrix::curv_scale().

double ndfboundp	(	double	x,
		double	fmin,
		double	fmax,
		const double &	fpen
	)

Derivatative code for double nd2fboundp( double x, double fmin, double fmax,const double& fpen)

Parameters

x	Variable to be scaled
fmin	Lower bound of x
fmin	Upper bound of x
fpen	Unchanged on return.

Returns

Derivative of scaled variable

Definition at line 210 of file boundfun.cpp.

Referenced by param_init_bounded_vector::mc_scale(), param_init_bounded_number::mc_scale(), param_init_bounded_matrix::mc_scale(), param_init_bounded_vector::sd_scale(), param_init_bounded_number::sd_scale(), param_init_bounded_matrix::sd_scale(), df1b2_init_bounded_number::sd_scale(), df1b2_init_bounded_vector::sd_scale(), and df1b2_init_bounded_matrix::sd_scale().