LWPR_Workspace Struct Reference

LWPR_Workspace is a structure that contains enough "working memory" for the LWPR computational routines. More...

#include <lwpr_aux.h>

Data Fields

int * derivOk
 Used within lwpr_aux_update_distance_metric for storing which PLS directions can be trusted.
double * storage
 Pointer to the allocated memory.
double * dx
 Used to hold the difference between a normalised input vector and a RF's centre.
double * dwdM
 Derivatives of the weight w with respect to LWPR_ReceptiveField.M.
double * dJ2dM
 Derivatives of the cost J2 with respect to M.
double * ddwdMdM
 2nd derivatives of w wrt. M
double * ddJ2dMdM
 2nd derivatives of J2 wrt. M
double * Ps
 Intermediate results used within lwpr_aux_update_distance_metric.
double * Pse
 Intermediate results used within lwpr_aux_update_distance_metric.
double * xu
 Used within PLS calculations (updated x).
double * yres
 Intermediate results used within lwpr_aux_update_regression.
double * ytarget
 Intermediate results used within lwpr_aux_update_regression.
double * xres
 Intermediate results used within lwpr_aux_update_regression.
double * xc
 Used to hold the difference between a normalised input vector and a RF's centre.
double * xmz
 Updated mean of a RF.
double * e_cv
 Intermediate results used within lwpr_aux_update_regression.
double * s
 Intermediate results used within lwpr_aux_update_regression.
double * dsdx
 Intermediate results used within lwpr_aux_predict_one_J.
double * Dx
 Used to store RF.D * (x-RF.c).
double * sum_dwdx
 Intermediate results used within lwpr_aux_predict_one_J.
double * sum_ydwdx_wdydx
 Intermediate results used within lwpr_aux_predict_one_J.
double * sum_ddwdxdx
 Intermediate results used within lwpr_aux_predict_one_gH.
double * sum_ddRdxdx
 Intermediate results used within lwpr_aux_predict_one_gH.

Detailed Description

LWPR_Workspace is a structure that contains enough "working memory" for the LWPR computational routines.

It is automatically allocated within lwpr_init_model, with one LWPR_Workspace structure per thread. You should not have to handle any of its elements yourself.


The documentation for this struct was generated from the following file:
Generated on Thu Feb 9 08:18:36 2012 for LWPR Library by  doxygen 1.6.3