hype::core::SchedulingDecision Class Reference

This class represents a scheduling decision for an operation w.r.t. a user specified set of features of the input data. More...

#include <scheduling_decision.hpp>

Collaboration diagram for hype::core::SchedulingDecision:

List of all members.

Public Member Functions
	SchedulingDecision (Algorithm &alg_ref, const EstimatedTime &estimated_time_for_algorithm, const Tuple &feature_values)
	constructs a SchedulingDecision object by assigning neccessary informations to all fields of the object
	~SchedulingDecision ()
const std::string	getNameofChoosenAlgorithm () const
	returns the name of the choosen algorithm
const EstimatedTime	getEstimatedExecutionTimeforAlgorithm () const
	returns the estimated execution time of the choosen algorithm
const Tuple	getFeatureValues () const
	returns the feature values that were the basis for this decision
const DeviceSpecification	getDeviceSpecification () const throw ()
	returns the ComputeDevice the chosen algorithm utilizes
bool	operator== (const SchedulingDecision &sched_dec) const
Private Attributes
Algorithm &	alg_ref_
	name of choosen algorithm
EstimatedTime	estimated_time_for_algorithm_
	estimated execution time of the choosen algorithm
Tuple	feature_values_
	feature values of input data set
uint64_t	scheduling_id_
	unique id which is unique for each schduling decision

---

Detailed Description

This class represents a scheduling decision for an operation w.r.t. a user specified set of features of the input data.

A user has to execute the algortihm suggested by the SchedulingDecision, or the library will not work. The user can determine the algorithm to excute by calling getNameofChoosenAlgorithm(). The user can then measure the execution time of the algorithm by using an AlgorithmMeasurement object.

Author:

Sebastian Breß

Version:

0.1

Date:

2012

Todo:

change interface, so it stores a pointer to the choosen algorithm, and update algorithm timestamps in the constructor!

Copyright:

GNU LESSER GENERAL PUBLIC LICENSE - Version 3, http://www.gnu.org/licenses/lgpl-3.0.txt

---

Constructor & Destructor Documentation

hype::core::SchedulingDecision::SchedulingDecision	(	Algorithm &	alg_ref,
		const EstimatedTime &	estimated_time_for_algorithm,
		const Tuple &	feature_values
	)

constructs a SchedulingDecision object by assigning neccessary informations to all fields of the object

Parameters:

alg_ref	Reference to Algorithm that was choosen
estimated_time_for_algorithm	estimated execution time for algorithm
feature_values	features of the input data set that were previously specified by the user

hype::core::SchedulingDecision::~SchedulingDecision

(

)

---

Member Function Documentation

const DeviceSpecification hype::core::SchedulingDecision::getDeviceSpecification

(

)

const throw ()

returns the ComputeDevice the chosen algorithm utilizes

Returns:

ComputeDevice the chosen algorithm utilizes

const EstimatedTime hype::core::SchedulingDecision::getEstimatedExecutionTimeforAlgorithm

(

)

const

returns the estimated execution time of the choosen algorithm

Returns:

estimated execution time of the choosen algorithm

const Tuple hype::core::SchedulingDecision::getFeatureValues

(

)

const

returns the feature values that were the basis for this decision

Returns:

feature values of input data set

const std::string hype::core::SchedulingDecision::getNameofChoosenAlgorithm

(

)

const

returns the name of the choosen algorithm

Returns:

name of choosen algorithm

Referenced by hype::queryprocessing::OperatorMapper< TablePtr, function >::getPhysicalOperator().

Here is the caller graph for this function:

bool hype::core::SchedulingDecision::operator==

(

const SchedulingDecision &

sched_dec

)

const

---

Member Data Documentation

Algorithm& hype::core::SchedulingDecision::alg_ref_ [private]

name of choosen algorithm

reference to algorithm

EstimatedTime hype::core::SchedulingDecision::estimated_time_for_algorithm_ [private]

estimated execution time of the choosen algorithm

Tuple hype::core::SchedulingDecision::feature_values_ [private]

feature values of input data set

uint64_t hype::core::SchedulingDecision::scheduling_id_ [private]

unique id which is unique for each schduling decision

---

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

• /home/sebastian/gpudbms/trunk/hype-library/include/core/scheduling_decision.hpp