Main Page   Class Hierarchy   Compound List   File List   Compound Members  

GlareAlgorithmGoodHitFactory Class Reference

#include <GlareAlgorithmGoodHitFactory.h>

List of all members.

Static Public Methods

GlareAlgorithmbuild (std::vector< Library * > p_libraries, std::vector< LibrarySet * > p_library_sets, GoodnessRule p_goodness_rule, float p_fraction_good_hit_threshold, int p_step_number_limit, bool p_do_scale_with_size, float p_scale_with_size_param, bool p_cut_zero_occurence, std::string p_enumeration_strategy, float p_first_fraction_to_keep, int p_minimum_partition_size)


Detailed Description

Class that instantiate and initialize a GlareAlgorithmGoodHit object.


Member Function Documentation

GlareAlgorithm* GlareAlgorithmGoodHitFactory::build std::vector< Library * >    p_libraries,
std::vector< LibrarySet * >    p_library_sets,
GoodnessRule    p_goodness_rule,
float    p_fraction_good_hit_threshold,
int    p_step_number_limit,
bool    p_do_scale_with_size,
float    p_scale_with_size_param,
bool    p_cut_zero_occurence,
std::string    p_enumeration_strategy,
float    p_first_fraction_to_keep,
int    p_minimum_partition_size
[static]
 

Build an abstract GlareAlgorithm object given a list of data built from user input data.

Parameters:
p_libraries  List of Library objects used in the optimization process.
p_library_sets  List of LibrarySet objects to be used in producing the combinatorial product. It must contain only Library objects also found in p_libraries parameter.
p_goodness_rule  Object containing the minimum and maximum value threshold for each property to be analyzed.
p_fraction_good_hit_threshold  Identifies the goodness that is sufficient to stop the optimization.
p_step_number_limit  Maximum number of iteration to be performed. The optimization process stops after that number of iterations.
p_do_scale_with_size  If set to 'true' apply the scaled pruning technique.
p_scale_with_size_param  When doing scaled pruning, this is the exponent used that defined the steepness of the sigmoid function.
p_cut_zero_occurence  If set to 'true' only the Fragment with a zero occurence count are eliminated after the first pass. This is usually inefficient.
p_enumeration_strategy  If set to 'part' partitioning scheme is expected to be used.
p_first_fraction_to_keep  Is the initial fraction of best Fragment to keep after the first iteration. 1-p_first_fraction_to_keep Fragment will be rejected. This corresponds to K0. A negative value automatically sets for the best optimal value as determined with a published linear equation.
p_minimum_partition_size  Is the smallest number of Fragment contained in a subdivided (partitioned) Library. Relevant only if partitioning scheme is used.


The documentation for this class was generated from the following file:
Generated on Sat Nov 4 15:59:00 2006 for GLARE by doxygen1.2.18