---

lev_marq.h File Reference

#include <gandalf/linalg/vec_gen.h>
#include <gandalf/linalg/mat_gen.h>
#include <gandalf/linalg/mat_square.h>
#include <gandalf/common/linked_list.h>

Go to the source code of this file.

Compounds
struct	Gan_LevMarqObs
	Observation structure for Levenberg-Marquardt minimisation. More...
struct	Gan_LevMarqStruct
	Structure for holding state of Levenberg-Marquardt algorithm. More...
Typedefs
typedef Gan_Bool(*	Gan_LevMarqObsFunc_h )(Gan_Vector *x, Gan_Vector *z, void *zdata, Gan_Vector *h, Gan_Matrix *H)
	Observation function for standard h-type observations.
typedef Gan_Bool(*	Gan_LevMarqObsFunc_F )(Gan_Vector *x, Gan_Vector *z, void *zdata, Gan_Vector *F, Gan_Matrix *Hx, Gan_Matrix *Hz)
	Observation function for F-type observations.
typedef Gan_LevMarqObs	Gan_LevMarqObs
	Observation structure for Levenberg-Marquardt minimisation.
typedef Gan_LevMarqStruct	Gan_LevMarqStruct
	Structure for holding state of Levenberg-Marquardt algorithm.
typedef Gan_Bool(*	Gan_LevMarqInitFunc )(Gan_Vector *x, Gan_List *obs_list, void *data)
	Callback function for initialising Levenberg-Marquardt algorithm.
Enumerations
enum	Gan_LevMarqObsType { GAN_LEV_MARQ_OBS_H, GAN_LEV_MARQ_OBS_H_ROBUST, GAN_LEV_MARQ_OBS_F }
	Observation type for Levenberg-Marquardt minimisation. More...
Functions
Gan_LevMarqStruct *	gan_lev_marq_form (Gan_LevMarqStruct *lm)
	Forms a Levenberg-Marquardt structure.
Gan_LevMarqObs *	gan_lev_marq_obs_h (Gan_LevMarqStruct *lm, Gan_Vector *z, void *zdata, Gan_SquMatrix *Ni, Gan_LevMarqObsFunc_h obs_func)
	Passes an observation to a Levenberg-Marquardt structure.
Gan_LevMarqObs *	gan_lev_marq_obs_h_robust (Gan_LevMarqStruct *lm, Gan_Vector *z, void *zdata, Gan_SquMatrix *Ni, Gan_LevMarqObsFunc_h obs_func, double var_scale, double chi2)
	Passes a robust observation to a Levenberg-Marquardt structure.
Gan_LevMarqObs *	gan_lev_marq_obs_F (Gan_LevMarqStruct *lm, Gan_Vector *z, void *zdata, Gan_SquMatrix *Ni, Gan_LevMarqObsFunc_F obs_func)
	Passes an observation to a Levenberg-Marquardt structure.
Gan_Bool	gan_lev_marq_init (Gan_LevMarqStruct *lm, Gan_LevMarqInitFunc init_func, void *data, double *residualp)
	Initialise Levenberg-Marquardt algorithm.
Gan_Bool	gan_lev_marq_iteration (Gan_LevMarqStruct *lm, double lambda, double *residual)
	Applies Levenberg-Marquardt iteration.
Gan_Vector *	gan_lev_marq_get_x (Gan_LevMarqStruct *lm)
	Returns state of Levenberg-Marquardt minimisation.
Gan_SquMatrix *	gan_lev_marq_get_P (Gan_LevMarqStruct *lm)
	Returns state covariance of Levenberg-Marquardt minimisation.
void	gan_lev_marq_free (Gan_LevMarqStruct *lm)
	Frees a Levenberg-Marquardt structure.
Gan_LevMarqStruct *	gan_lev_marq_alloc (void)
	Macro: Allocates a Levenberg-Marquardt structure.

---

Detailed Description

Module: Levenberg-Marquardt non-linear minimisation algorithm

Part of: Gandalf Library

Revision: Last edited: Author:

Copyright: (c) 2000 Imagineer Software Limited

---

Function Documentation

Gan_LevMarqStruct* gan_lev_marq_form (  Gan_LevMarqStruct *    lm )

Forms a Levenberg-Marquardt structure. Parameters: lm  Pointer to a Levenberg-Marquardt structure or NULL Allocates/fills a structure for computing Levenberg-Marquardt optimisation. Returns: Pointer to structure or NULL on failure. See also: gan_lev_marq_alloc().

void gan_lev_marq_free (  Gan_LevMarqStruct *    lm )

Frees a Levenberg-Marquardt structure. Parameters: lm  Pointer to Levenberg-Marquardt structure Frees a structure created to compute Levenberg-Marquardt minimisation. Returns: None See also: gan_lev_marq_form().

Gan_SquMatrix* gan_lev_marq_get_P (  Gan_LevMarqStruct *    lm )

Returns state covariance of Levenberg-Marquardt minimisation. Parameters: lm  Pointer to a Levenberg-Marquardt structure Returns the state covariance of a Levenberg-Marquardt minimisation. The matrix is not copied, but is rather passed back directly from the state structure. There is therefore no need to free the matrix after use. It is essential also that the contents of the matrix are not modified; i.e. the matrix should be treated as read only. Returns: Pointer to current solution covariance or NULL on failure. See also: gan_lev_marq_get_x().

Gan_Vector* gan_lev_marq_get_x (  Gan_LevMarqStruct *    lm )

Returns state of Levenberg-Marquardt minimisation. Parameters: lm  Pointer to a Levenberg-Marquardt structure Returns the state of a Levenberg-Marquardt minimisation. The vector is not copied, but is rather passed back directly from the state structure. There is therefore no need to free the vector after use. It is essential also that the contents of the vector are not modified; i.e. the vector should be treated as read only. Returns: Pointer to current solution vector or NULL on failure. See also: gan_lev_marq_get_P().

Gan_Bool gan_lev_marq_init (  Gan_LevMarqStruct *    lm, Gan_LevMarqInitFunc    init_func, void *    data, double *    residualp )

Initialise Levenberg-Marquardt algorithm. Parameters: lm  Pointer to a Levenberg-Marquardt structure init_func  Function to initialise value of state vector data  User pointer to be passed into init_func residualp  Pointer to initial residual After passing some observations to the structure using gan_lev_marq_obs_h() or gan_lev_marq_obs_F(), this function invokes the callback function #init_func to set the state to initial values, before applying Levenberg-Marquardt iterations using gan_lev_marq_iteration(). Returns: GAN_TRUE on successful initialisation of structure, GAN_FALSE on failure. See also: gan_lev_marq_iteration().

Gan_Bool gan_lev_marq_iteration (  Gan_LevMarqStruct *    lm, double    lambda, double *    residualp )

Applies Levenberg-Marquardt iteration. Parameters: lm  Pointer to a Levenberg-Marquardt structure lambda  Adjustment factor for adjusting the information matrix residualp  Output residual after iteration Applies a single Levenberg-Marquardt iteration. The residual after the adjustment is returned in the provided pointer. The adjustment is only actually applied to the state vector if the residual has been reduced from the existing value. Returns: GAN_TRUE on successful completion of iteration, GAN_FALSE on failure.

Gan_LevMarqObs* gan_lev_marq_obs_F (  Gan_LevMarqStruct *    lm, Gan_Vector *    z, void *    zdata, Gan_SquMatrix *    Ni, Gan_LevMarqObsFunc_F    obs_func )

Passes an observation to a Levenberg-Marquardt structure. Parameters: lm  Pointer to a Levenberg-Marquardt structure z  Observation vector zdata  Data pointer to pass along with z Ni  Observation inverse covariance obs_func  Observation callback function Adds a new observation for Levenberg-Marquardt minimisation. It relates state to observation according to the model where is the observation vector, is the observation function, is the state vector and is a zero-mean noise vector, with covariance . Returns: Handle for the new observation or NULL on failure. See also: gan_lev_marq_form().

Gan_LevMarqObs* gan_lev_marq_obs_h (  Gan_LevMarqStruct *    lm, Gan_Vector *    z, void *    zdata, Gan_SquMatrix *    Ni, Gan_LevMarqObsFunc_h    obs_func )

Passes an observation to a Levenberg-Marquardt structure. Parameters: lm  Pointer to a structure z  Observation vector zdata  Data pointer to pass along with z Ni  Observation inverse covariance obs_func  Observation callback function Adds a new observation for Levenberg-Marquardt minimisation. The observation is a direct observation relating state to observation where is the observation vector, is the observation function, is the state vector and is a zero-mean noise vector, with covariance . Returns: Handle for the new observation or NULL on failure. See also: gan_lev_marq_form().

Gan_LevMarqObs* gan_lev_marq_obs_h_robust (  Gan_LevMarqStruct *    lm, Gan_Vector *    z, void *    zdata, Gan_SquMatrix *    Ni, Gan_LevMarqObsFunc_h    obs_func, double    var_scale, double    chi2 )

Passes a robust observation to a Levenberg-Marquardt structure. Parameters: lm  Pointer to a structure z  Observation vector zdata  Data pointer to pass along with z Ni  Observation inverse covariance obs_func  Observation callback function var_scale  Scaling of outlier covariance w.r.t. inlier covariance chi2  Cutoff point on innovation for switching distributions Adds a new observation for Levenberg-Marquardt minimisation. The observation is a direct observation relating state to observation where is the observation vector, is the observation function, is the state vector and is a zero-mean noise vector, with covariance . Returns: Handle for the new observation or NULL on failure. See also: gan_lev_marq_form().

---