Sampling_check Class Reference

Class to check an existing sample set, if a constraint is fullfilled. More...

#include <sampling.h>

Collaboration diagram for Sampling_check:
Collaboration graph
[legend]

List of all members.

Public Member Functions

 Sampling_check (Pointer< Param > param_)
 (Default-)Constructor.
void check (vector< vector< dvector > > &sample_set, const SepQcFunc &f, bool eq, const vector< Pointer< set< int > > > &i_lin, const UserVector< double > &lower, const UserVector< double > &upper, const vector< int > &start, int min_keep=0)
 Computes sample points by sampling in one dimension less and adjusting the linear variable, such that a constraint is fullfilled.
void check (vector< vector< dvector > > &sample_set, const SepQcFunc &f, bool eq, const UserVector< double > &lower, const UserVector< double > &upper, const vector< int > &start, int min_keep)

Private Attributes

Pointer< Paramparam

Detailed Description

Class to check an existing sample set, if a constraint is fullfilled.

If the constraint is not fullfilled in a point, the point is tried to make feasible by adjusting the linear variables.

Definition at line 216 of file sampling.h.


Constructor & Destructor Documentation

Sampling_check::Sampling_check ( Pointer< Param param_  )  [inline]

(Default-)Constructor.

Definition at line 223 of file sampling.h.


Member Function Documentation

void Sampling_check::check ( vector< vector< dvector > > &  sample_set,
const SepQcFunc f,
bool  eq,
const vector< Pointer< set< int > > > &  i_lin,
const UserVector< double > &  lower,
const UserVector< double > &  upper,
const vector< int > &  start,
int  min_keep = 0 
)

Computes sample points by sampling in one dimension less and adjusting the linear variable, such that a constraint is fullfilled.

If j is a linear variable, samples in {1,..,j-1,j+1,..}. If x_j can be set, such that f(x)==0 or f(x)<=0 and x_j is in its bounds, the point is added to the sample set. If x_j cannot set this way, it's set to on of its bounds and the next variable is tried. This method prints a warning, if no points were added. If no linear variables exists, it generates a normal sample set.

Parameters:
sample_set The sample sets for each block of f to add the points to.
f The function for the constraint f(x) == 0 or f(x) <= 0.
eq Indicates, whether the function is an equation or inequation.
i_lin The indices of the linear variables for each block.
lower The lower bounds of the variables.
upper The upper bounds of the variables
start The first sample point in the sample set, we should check.
min_keep The minimum number of sample points to keep in the sample set, if somehow possible.
void Sampling_check::check ( vector< vector< dvector > > &  sample_set,
const SepQcFunc f,
bool  eq,
const UserVector< double > &  lower,
const UserVector< double > &  upper,
const vector< int > &  start,
int  min_keep 
)

Member Data Documentation

Definition at line 218 of file sampling.h.


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

Generated on 10 Mar 2013 for LaGO by  doxygen 1.6.1