Implements the functions needed to perform model averaging over the n best Bayesian networks. More...

#include "model_averaging.h"
#include "parent_set_data.h"

Include dependency graph for model_averaging.c:

Functions
static int	indexOf (MA_info ma_info, SCIP_VAR var)
	Find the index of ma_info->vars and ma_info->average_scores relating to a variable. More...

SCIP_RETCODE	MA_addAveragingParameters (SCIP scip, MA_info ma_info)
	Adds parameters for controlling the model averaging. More...

SCIP_RETCODE	MA_createAverageDataStructure (SCIP scip, MA_info ma_info)
	Allocates memory for the data structures used for model averaging. More...

SCIP_RETCODE	MA_destroyAverageDataStructure (SCIP scip, MA_info ma_info)
	Frees memory used for the data structures used for model averaging. More...

SCIP_Real	MA_getAverageValue (MA_info ma_info, SCIP_VAR variable)
	Returns the model average value of a given variable. More...

SCIP_Real	MA_getTotalAveragesScore (MA_info *ma_info)
	Returns the total likelihood of all the solutions included in the model average. More...

SCIP_Real	MA_getTotalAveragesTime (MA_info *ma_info)
	Returns the total time spent solving for all the solutions included in the model average. More...

SCIP_RETCODE	MA_updateAverageDataStructure (SCIP scip, MA_info ma_info, SCIP_SOL *sol)
	Updates the average scores based on a newly found solution. More...

Detailed Description

Implements the functions needed to perform model averaging over the n best Bayesian networks.

Function Documentation

◆ indexOf()

static int indexOf	(	MA_info *	ma_info,
		SCIP_VAR *	var
	)

static

Find the index of ma_info->vars and ma_info->average_scores relating to a variable.

Parameters

var	The variable of interest.
ma_info	Model averaging information.

Returns: The index of the var in ma_info->vars. If the variable is not included in the averaging, -1 is returned.

References MA_info::num_vars, and MA_info::vars.

Referenced by MA_getAverageValue().

◆ MA_addAveragingParameters()

SCIP_RETCODE MA_addAveragingParameters	(	SCIP *	scip,
		MA_info *	ma_info
	)

Adds parameters for controlling the model averaging.

Parameters

scip	The SCIP instance to which the parameter is to be added.
ma_info	Model averaging information.

Returns: SCIP_OKAY if the parameters were added successfully or an error code otherwise.

References MA_info::is_log_score.

Referenced by main().

◆ MA_createAverageDataStructure()

SCIP_RETCODE MA_createAverageDataStructure	(	SCIP *	scip,
		MA_info *	ma_info
	)

Allocates memory for the data structures used for model averaging.

Parameters

scip	The SCIP instance on which the model averaging will be performed.
ma_info	Model averaging information.

Returns: SCIP_OKAY if memory allocation was successful or an appropriate error message otherwise.

References MA_info::average_scores, MA_info::first_score, MA_info::is_first_score, MA_info::normalising_constant, MA_info::num_vars, MA_info::total_score, MA_info::total_time, and MA_info::vars.

Referenced by main().

◆ MA_destroyAverageDataStructure()

SCIP_RETCODE MA_destroyAverageDataStructure	(	SCIP *	scip,
		MA_info *	ma_info
	)

Frees memory used for the data structures used for model averaging.

Parameters

scip	The SCIP instance on which the model averaging was performed.
ma_info	Model averaging information.

Returns: SCIP_OKAY if memory deallocation was successful or an appropriate error message otherwise.

References MA_info::average_scores, and MA_info::vars.

Referenced by main().

◆ MA_getAverageValue()

SCIP_Real MA_getAverageValue	(	MA_info *	ma_info,
		SCIP_VAR *	variable
	)

Returns the model average value of a given variable.

If the variable is not part of this model averaging, the value -1 will be returned.

Parameters

ma_info	Model averaging information.
variable	The variable to get the model averaging score for.

Returns: The model average score of the variable.

References MA_info::average_scores, indexOf(), MA_info::is_log_score, MA_info::normalising_constant, and MA_info::total_score.

Referenced by printAverages().

Here is the call graph for this function:

◆ MA_getTotalAveragesScore()

SCIP_Real MA_getTotalAveragesScore ( MA_info * ma_info )

Returns the total likelihood of all the solutions included in the model average.

Parameters

ma_info Model averaging information.

Returns: The sum of the likelihoods of each of the solutions included.

References MA_info::total_score.

Referenced by printAverages().

◆ MA_getTotalAveragesTime()

SCIP_Real MA_getTotalAveragesTime ( MA_info * ma_info )

Returns the total time spent solving for all the solutions included in the model average.

Parameters

ma_info Model averaging information.

Returns: The number of seconds spend on solving all of the solutions that have been used to make up the average.

References MA_info::total_time.

Referenced by printAverages().

◆ MA_updateAverageDataStructure()

SCIP_RETCODE MA_updateAverageDataStructure	(	SCIP *	scip,
		MA_info *	ma_info,
		SCIP_SOL *	sol
	)

Updates the average scores based on a newly found solution.

Parameters

scip	The SCIP instance to which the solution belongs.
ma_info	Model averaging information.
sol	The new solution to incorporate in to the averages.

Returns: SCIP_OKAY if the operation succeeded or an appropriate error message otherwise.

References MA_info::average_scores, MA_info::first_score, MA_info::is_first_score, MA_info::is_log_score, MA_info::normalising_constant, MA_info::num_vars, MA_info::total_score, MA_info::total_time, and MA_info::vars.

Referenced by main().

Functions

Detailed Description

Function Documentation

◆ indexOf()

◆ MA_addAveragingParameters()

◆ MA_createAverageDataStructure()

◆ MA_destroyAverageDataStructure()

◆ MA_getAverageValue()

◆ MA_getTotalAveragesScore()

◆ MA_getTotalAveragesTime()

◆ MA_updateAverageDataStructure()