RTOp_ROp_max_step.h File Reference

#include "RTOp.h"
Include dependency graph for RTOp_ROp_max_step.h:

Go to the source code of this file.

Reduction operator for finding the maximum step for feasibility.

targ_obj <- { max alpha | v[0] + alpha * v[1] >= beta }

This is a specialized reduction operation that is used in many optimization methods to find the maximum step length alpha such that the iterates remain feasible. It is assumed that v[0] > beta so that alpha == 0.0 is a valid return. If the step is unrestricted, the function RTOp_ROp_max_step_val() will return RTOp_ROp_max_step_inf.

This operator is defined to allow exactly two vector arguments (num_vecs == 2) v[0], v[1], and can only handle dense vectors.

struct RTOp_RTOp_vtbl_t RTOp_ROp_max_step_vtbl
RTOp_value_type RTOp_ROp_max_step_inf
int RTOp_ROp_max_step_construct (RTOp_value_type beta, struct RTOp_RTOp *op)
int RTOp_ROp_max_step_destroy (struct RTOp_RTOp *op)
int RTOp_ROp_max_step_set_beta (RTOp_value_type beta, struct RTOp_RTOp *op)
RTOp_value_type RTOp_ROp_max_step_val (RTOp_ReductTarget reduct_obj)

Function Documentation

int RTOp_ROp_max_step_construct ( RTOp_value_type  beta,
struct RTOp_RTOp op 

Definition at line 90 of file RTOp_ROp_max_step.c.

int RTOp_ROp_max_step_destroy ( struct RTOp_RTOp op  ) 

Definition at line 98 of file RTOp_ROp_max_step.c.

int RTOp_ROp_max_step_set_beta ( RTOp_value_type  beta,
struct RTOp_RTOp op 

Definition at line 105 of file RTOp_ROp_max_step.c.

RTOp_value_type RTOp_ROp_max_step_val ( RTOp_ReductTarget  reduct_obj  ) 

Definition at line 113 of file RTOp_ROp_max_step.c.

Variable Documentation

Definition at line 77 of file RTOp_ROp_max_step.c.

Definition at line 111 of file RTOp_ROp_max_step.c.

 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines
Generated on Wed Apr 13 10:16:53 2011 for MOOCHO (Single Doxygen Collection) by  doxygen 1.6.3