hype::core::Operation Class Reference

This class represents an operation, which abstracts from algorithms and summarizes them under an abstract task, like sorting, join, selection, projection etc. More...

#include <operation.hpp>

Collaboration diagram for hype::core::Operation:

List of all members.

Public Types
typedef std::map< std::string, std::tr1::shared_ptr < Algorithm > >	MapNameToAlgorithm
Public Member Functions
	Operation (const std::string &name)
	~Operation ()
bool	addAlgorithm (const std::string &name_of_algorithm, DeviceSpecification comp_dev, const std::string &name_of_statistical_method, const std::string &name_of_recomputation_strategy)
void	removeAlgorithm (const std::string &name)
const SchedulingDecision	getOptimalAlgorithm (const Tuple &input_values, DeviceTypeConstraint dev_constr=ANY_DEVICE)
bool	hasAlgorithm (const std::string &name_of_algorithm)
	returns true if an algrithm named "name_of_algorithm" is registered for this operation and false otherwise
const AlgorithmPtr	getAlgorithm (const std::string &name_of_algorithm)
bool	setNewOptimizationCriterion (const std::string &name_of_optimization_criterion)
bool	addObservation (const std::string &name_of_algorithm, const MeasurementPair &mp)
const std::vector< AlgorithmPtr >	getAlgorithms ()
const std::map< double, std::string >	getEstimatedExecutionTimesforAlgorithms (const Tuple &input_values)
const std::string	getName () const throw ()
void	incrementNumberOfRightDecisions () throw ()
void	incrementNumberOfTotalDecisions () throw ()
uint64_t	getNextTimestamp () throw ()
uint64_t	getCurrentTimestamp () const throw ()
Private Attributes
MapNameToAlgorithm	map_algorithmname_to_pointer_
std::tr1::shared_ptr < OptimizationCriterion_Internal >	ptr_to_optimization_criterion_
std::string	name_
unsigned int	number_of_right_decisions_
unsigned int	number_of_total_decisions_
uint64_t	logical_time_

---

Detailed Description

This class represents an operation, which abstracts from algorithms and summarizes them under an abstract task, like sorting, join, selection, projection etc.

An operation gets assigned two things. First is a list of algorithms, which represent the set of algorithms available to process operation O, which is basically the algorithm pool of the decision model. Second is an optimization_criterion, which is implemented by drived classes of stemod::core::OptimizationCriterion. The optimization criterion takes the computed estimated execution times of each algorithm and decides on the optimal algorithm, where the user specifies what is optimal by specifying an OptimizationCriterion.

Note that the stemod::core::OptimizationCriterioc can be exchanged at run-time, because the Operation class uses the pointer to implementation technique (or pimpl-idiom). This makes it highly adaptable.

Author:

Sebastian Breß

Version:

0.1

Date:

2012

Copyright:

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

---

Member Typedef Documentation

typedef std::map<std::string,std::tr1::shared_ptr<Algorithm> > hype::core::Operation::MapNameToAlgorithm

---

Constructor & Destructor Documentation

hype::core::Operation::Operation

(

const std::string &

name

)

hype::core::Operation::~Operation

(

)

---

Member Function Documentation

bool hype::core::Operation::addAlgorithm	(	const std::string &	name_of_algorithm,
		DeviceSpecification	comp_dev,
		const std::string &	name_of_statistical_method,
		const std::string &	name_of_recomputation_strategy
	)

bool hype::core::Operation::addObservation	(	const std::string &	name_of_algorithm,
		const MeasurementPair &	mp
	)

const AlgorithmPtr hype::core::Operation::getAlgorithm

(

const std::string &

name_of_algorithm

)

const std::vector<AlgorithmPtr> hype::core::Operation::getAlgorithms

(

)

uint64_t hype::core::Operation::getCurrentTimestamp

(

)

const throw ()

const std::map<double,std::string> hype::core::Operation::getEstimatedExecutionTimesforAlgorithms

(

const Tuple &

input_values

)

const std::string hype::core::Operation::getName

(

)

const throw ()

uint64_t hype::core::Operation::getNextTimestamp

(

)

throw ()

const SchedulingDecision hype::core::Operation::getOptimalAlgorithm	(	const Tuple &	input_values,
		DeviceTypeConstraint	dev_constr = ANY_DEVICE
	)

bool hype::core::Operation::hasAlgorithm

(

const std::string &

name_of_algorithm

)

returns true if an algrithm named "name_of_algorithm" is registered for this operation and false otherwise

void hype::core::Operation::incrementNumberOfRightDecisions

(

)

throw ()

void hype::core::Operation::incrementNumberOfTotalDecisions

(

)

throw ()

void hype::core::Operation::removeAlgorithm

(

const std::string &

name

)

bool hype::core::Operation::setNewOptimizationCriterion

(

const std::string &

name_of_optimization_criterion

)

---

Member Data Documentation

uint64_t hype::core::Operation::logical_time_ [private]

MapNameToAlgorithm hype::core::Operation::map_algorithmname_to_pointer_ [private]

std::string hype::core::Operation::name_ [private]

unsigned int hype::core::Operation::number_of_right_decisions_ [private]

unsigned int hype::core::Operation::number_of_total_decisions_ [private]

std::tr1::shared_ptr<OptimizationCriterion_Internal> hype::core::Operation::ptr_to_optimization_criterion_ [private]

---

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

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