def __init__(self, particle):
""" The Topology Class Creator """
print "New topology instance, %s class particles." % (
particle.__class__.__name__, )
self.oneSelfParticle = particle
self.internalSwarm = []
self.swarmSize = 0
self.sortType = Consts.CDefSwarmSortType
self.sorted = False
self.minimax = Consts.CDefSwarmMinimax
#Best particle inside topology
self.bestParticle = None
self.position_updater = FunctionSlot("Position Particles Updater")
self.information_updater = FunctionSlot(
"Information Particles Updater")
self.allSlots = [self.position_updater, self.information_updater]
#Statistics
self.statted = False
self.topologyStats = TopologyStatistics()
self.swarmStats = SwarmStatistics()
def __init__(self,particle):
""" The Topology Class Creator """
print "New topology instance, %s class particles." % (particle.__class__.__name__,)
self.oneSelfParticle = particle
self.internalSwarm = []
self.swarmSize = 0
self.sortType = Consts.CDefSwarmSortType
self.sorted = False
self.minimax = Consts.CDefSwarmMinimax
#Best particle inside topology
self.bestParticle = None
self.position_updater = FunctionSlot("Position Particles Updater")
self.information_updater = FunctionSlot("Information Particles Updater")
self.allSlots = [self.position_updater, self.information_updater]
#Statistics
self.statted = False
self.topologyStats = TopologyStatistics()
self.swarmStats = SwarmStatistics()
class TopologyBase(object):
""" Topology Base Class - The container for the swarm
**Examples**
Get the swarm from the :class:`PSO.SimplePSO` (PSO Engine) instance
>>> swarm = pso_engine.getSwarm()
Get the best fitness particle
>>> bestParticle = swarm.bestFitness()
Ge the statistics from the :class:`Statistics.Statistics` instance
>>> stats = swarm.getStatistics()
>>> print stats["fitMax"]
10.4
Iterate, get/set individuals
>>> for particle in swarm:
>>> print particle
(...)
>>> for i in xrange(len(swarm)):
>>> print swarm[i]
(...)
>>> swarm[10] = newParticle
>>> swarm[10].fitness
12.5
:param particle: the :term: `Sample particle``
"""
position_updater = None
""" This is the position update topology function slot, you can change the default
updater using the slot *set* function: ::
topology.position_updater.set(GlobalTopology.GlobalPositionUpdater)
"""
information_updater = None
""" This is the information update topology function slot, you can change the default
updater usingt the slot *set* function: ::
topology.information_communicator.set(GlobalTopology.GlobalInformationUpdater)
"""
def __init__(self, particle):
""" The Topology Class Creator """
print "New topology instance, %s class particles." % (
particle.__class__.__name__, )
self.oneSelfParticle = particle
self.internalSwarm = []
self.swarmSize = 0
self.sortType = Consts.CDefSwarmSortType
self.sorted = False
self.minimax = Consts.CDefSwarmMinimax
#Best particle inside topology
self.bestParticle = None
self.position_updater = FunctionSlot("Position Particles Updater")
self.information_updater = FunctionSlot(
"Information Particles Updater")
self.allSlots = [self.position_updater, self.information_updater]
#Statistics
self.statted = False
self.topologyStats = TopologyStatistics()
self.swarmStats = SwarmStatistics()
def setMinimax(self, minimax):
""" Sets the swarm minimax
Example:
>>> swarm.setMinimax(Consts.minimaxType["maximize"])
:param minimax: the minimax type
"""
self.minimax = minimax
def __repr__(self):
""" Returns the string representation of the topology """
ret = "- Topology\n"
ret += "\nSwarm Size:\t %d\n" % (self.swarmSize, )
ret += "Sort Type:\t\t %s\n" % (Consts.sortType.keys()[
Consts.sortType.values().index(self.sortType)].capitalize(), )
ret += "\tMinimax Type:\t\t %s\n" % (Consts.minimaxType.keys()[
Consts.minimaxType.values().index(self.minimax)].capitalize(), )
for slot in self.allSlots:
ret += "\t" + slot.__repr__()
ret += "\n"
ret += self.topologyStats.__repr__()
return ret
def setTopologyStatistics(self, topStats):
""" Sets the topology Statistics (Best Particle info)
Example:
>>> topology.setTopologyStatistics(FloatStatistics.TopologyStatistics())
:param topStats: the statistics instance class (must be extend Statistics base class)
"""
self.topologyStats = topStats
def setSwarmStatistics(self, swarmStats):
""" Sets the Swarm Statistics (Swarm info)
Example:
>>> topology.setSwarmStatistics(FloatStatistics.SwarmStatistics())
:param swarmStats: the statistics instance class (must be extend Statistics base class)
"""
self.swarmStats = swarmStats
def __len__(self):
""" Return the length of the swarm """
return len(self.internalSwarm)
def __getitem__(self, key):
""" Returns the specified particle from topology """
return self.internalSwarm[key]
def __iter__(self):
""" Returns the iterator of the swarm """
return iter(self.internalSwarm)
def __setitem__(self, key, value):
""" Set the particle of swarm """
self.internalSwarm[key] = value
self.__clear_flags()
def clear_flags(self):
self.statted = False
#self.statted = False
def statistics(self):
"""Do the statistical analysis of the swarm and set 'statted' to True """
if self.statted: return
fit_sum = 0.0
swarm_size = len(self.internalSwarm)
for index in xrange(swarm_size):
fit_sum += self.internalSwarm[index].fitness
self.swarmStats["fitMax"] = max(self.internalSwarm,
key=key_fitness_score).fitness
self.swarmStats["fitMin"] = min(self.internalSwarm,
key=key_fitness_score).fitness
self.swarmStats["fitAvg"] = fit_sum / float(swarm_size)
bestFit_sum = 0
for index in xrange(swarm_size):
bestFit_sum += self.internalSwarm[index].ownBestFitness
self.swarmStats["bestFitMin"] = min(
self.internalSwarm, key=key_best_fitness_score).ownBestFitness
self.swarmStats["bestFitMax"] = max(
self.internalSwarm, key=key_best_fitness_score).ownBestFitness
self.swarmStats["bestFitAvg"] = bestFit_sum / float(swarm_size)
tmpvar = 0.0
for index in xrange(swarm_size):
s = self.internalSwarm[index].ownBestFitness - self.swarmStats[
"bestFitAvg"]
s *= s
tmpvar += s
tmpvar /= float((len(self.internalSwarm) - 1))
self.swarmStats["bestFitVar"] = tmpvar
self.swarmStats["bestFitDev"] = math.sqrt(tmpvar)
self.topologyStats["bestFitness"] = self.bestParticle.ownBestFitness
self.topologyStats[
"bestPosition"] = self.bestParticle.ownBestPosition[:]
self.topologyStats["bestPosDim"] = self.bestParticle.ownBestPosition[0]
self.topologyStats["position"] = self.bestParticle.position[:]
self.topologyStats["fitness"] = self.bestParticle.fitness
self.statted = True
def getBestParticle(self):
""" Return the best particle of the swarm
:rtype: the particle
"""
return self.bestParticle
def sort(self):
""" Sort the swarm """
if self.sorted: return
rev = (self.minimax == Consts.minimaxType["maximize"])
if self.sortType == Consts.sortType["fitness"]:
self.internalSwarm.sort(cmp=cmp_particle_fitness, reverse=rev)
self.sorted = True
def setSwarmSize(self, size):
""" Set the population size
:param size: the population size
"""
self.swarmSize = size
def setSortType(self, sort_type):
""" Sets the sort type
Example:
>>> swarm.setSortType(Consts.sortType["scaled"])
:param sort_type: the Sort Type
"""
self.sortType = sort_type
def setBestParticle(self, particle):
""" Set the best Particle
Example:
>>> topology.setBestParticle(bestParticle)
:param particle: the best particle to set
"""
self.bestParticle = particle
def create(self, **args):
""" Clone the example particle fo fill the swarm """
self.clear()
self.minimax = args["minimax"]
for i in xrange(self.swarmSize):
self.internalSwarm.append(self.oneSelfParticle.clone())
self.clear_flags()
def initialize(self):
""" Initialize all particles of swarm
this calls the initialize() of particles
PS: You can OVERRIDE this method for custom initialization.
"""
for particle in self.internalSwarm:
particle.initializePosition()
particle.initializeVelocity()
particle.evaluate()
particle.ownBestFitness = particle.fitness
self.bestParticle = self.internalSwarm[0]
self.clear_flags()
def evaluate(self, *args):
""" Evaluate all particles in swarm, calls the evaluate() method of particles
:param args: this param are passed to the evaluation function
"""
for particle in self.internalSwarm:
particle.evaluate(**args)
self.clear_flags()
def printStats(self):
""" Print statistics of the current population """
message = ""
#message = "[Swarm] - Max/Min/Avg bestFitness(Fitness) [%.2f(%.2f)/%.2f(%.2f)/%.2f(%.2f)]\n" % (self.stats["bestFitMax"], self.stats["fitMax"], self.stats["fitMin"], self.stats["bestFitMin"], self.stats["bestFitAvg"], self.stats["fitAvg"])
message += "[Topology] - bestFitness/bestPosDim [%s/%s]" % (
self.topologyStats["bestFitness"],
self.topologyStats["bestPosDim"])
print message
return message
def clear(self):
""" Remove all particles from swarm """
del self.internalSwarm[:]
self.clear_flags()
def getStatistics(self):
""" Return a Statistics classes for statistics
:rtype: the subclasses of :class: `Statistics.Statistics` (topology,swarm) instances
"""
self.statistics()
self.clear_flags()
return (self.topologyStats, self.swarmStats)
class TopologyBase(object):
""" Topology Base Class - The container for the swarm
**Examples**
Get the swarm from the :class:`PSO.SimplePSO` (PSO Engine) instance
>>> swarm = pso_engine.getSwarm()
Get the best fitness particle
>>> bestParticle = swarm.bestFitness()
Ge the statistics from the :class:`Statistics.Statistics` instance
>>> stats = swarm.getStatistics()
>>> print stats["fitMax"]
10.4
Iterate, get/set individuals
>>> for particle in swarm:
>>> print particle
(...)
>>> for i in xrange(len(swarm)):
>>> print swarm[i]
(...)
>>> swarm[10] = newParticle
>>> swarm[10].fitness
12.5
:param particle: the :term: `Sample particle``
"""
position_updater = None
""" This is the position update topology function slot, you can change the default
updater using the slot *set* function: ::
topology.position_updater.set(GlobalTopology.GlobalPositionUpdater)
"""
information_updater = None
""" This is the information update topology function slot, you can change the default
updater usingt the slot *set* function: ::
topology.information_communicator.set(GlobalTopology.GlobalInformationUpdater)
"""
def __init__(self,particle):
""" The Topology Class Creator """
print "New topology instance, %s class particles." % (particle.__class__.__name__,)
self.oneSelfParticle = particle
self.internalSwarm = []
self.swarmSize = 0
self.sortType = Consts.CDefSwarmSortType
self.sorted = False
self.minimax = Consts.CDefSwarmMinimax
#Best particle inside topology
self.bestParticle = None
self.position_updater = FunctionSlot("Position Particles Updater")
self.information_updater = FunctionSlot("Information Particles Updater")
self.allSlots = [self.position_updater, self.information_updater]
#Statistics
self.statted = False
self.topologyStats = TopologyStatistics()
self.swarmStats = SwarmStatistics()
def setMinimax(self,minimax):
""" Sets the swarm minimax
Example:
>>> swarm.setMinimax(Consts.minimaxType["maximize"])
:param minimax: the minimax type
"""
self.minimax = minimax
def __repr__(self):
""" Returns the string representation of the topology """
ret = "- Topology\n"
ret += "\nSwarm Size:\t %d\n" % (self.swarmSize,)
ret += "Sort Type:\t\t %s\n" % (Consts.sortType.keys()[Consts.sortType.values().index(self.sortType)].capitalize(),)
ret += "\tMinimax Type:\t\t %s\n" % (Consts.minimaxType.keys()[Consts.minimaxType.values().index(self.minimax)].capitalize(),)
for slot in self.allSlots:
ret+= "\t" + slot.__repr__()
ret += "\n"
ret+= self.topologyStats.__repr__()
return ret
def setTopologyStatistics(self,topStats):
""" Sets the topology Statistics (Best Particle info)
Example:
>>> topology.setTopologyStatistics(FloatStatistics.TopologyStatistics())
:param topStats: the statistics instance class (must be extend Statistics base class)
"""
self.topologyStats = topStats
def setSwarmStatistics(self,swarmStats):
""" Sets the Swarm Statistics (Swarm info)
Example:
>>> topology.setSwarmStatistics(FloatStatistics.SwarmStatistics())
:param swarmStats: the statistics instance class (must be extend Statistics base class)
"""
self.swarmStats = swarmStats
def __len__(self):
""" Return the length of the swarm """
return len(self.internalSwarm)
def __getitem__(self, key):
""" Returns the specified particle from topology """
return self.internalSwarm[key]
def __iter__(self):
""" Returns the iterator of the swarm """
return iter(self.internalSwarm)
def __setitem__(self,key,value):
""" Set the particle of swarm """
self.internalSwarm[key] = value
self.__clear_flags()
def clear_flags(self):
self.statted = False
#self.statted = False
def statistics(self):
"""Do the statistical analysis of the swarm and set 'statted' to True """
if self.statted: return
fit_sum = 0.0
swarm_size = len(self.internalSwarm)
for index in xrange(swarm_size):
fit_sum += self.internalSwarm[index].fitness
self.swarmStats["fitMax"] = max(self.internalSwarm,key=key_fitness_score).fitness
self.swarmStats["fitMin"] = min(self.internalSwarm,key=key_fitness_score).fitness
self.swarmStats["fitAvg"] = fit_sum / float(swarm_size)
bestFit_sum = 0
for index in xrange(swarm_size):
bestFit_sum += self.internalSwarm[index].ownBestFitness
self.swarmStats["bestFitMin"] = min(self.internalSwarm,key=key_best_fitness_score).ownBestFitness
self.swarmStats["bestFitMax"] = max(self.internalSwarm,key=key_best_fitness_score).ownBestFitness
self.swarmStats["bestFitAvg"] = bestFit_sum / float(swarm_size)
tmpvar = 0.0
for index in xrange(swarm_size):
s = self.internalSwarm[index].ownBestFitness - self.swarmStats["bestFitAvg"]
s*=s
tmpvar += s
tmpvar /= float((len(self.internalSwarm)-1))
self.swarmStats["bestFitVar"] = tmpvar
self.swarmStats["bestFitDev"] = math.sqrt(tmpvar)
self.topologyStats["bestFitness"] = self.bestParticle.ownBestFitness
self.topologyStats["bestPosition"] = self.bestParticle.ownBestPosition[:]
self.topologyStats["bestPosDim"] = self.bestParticle.ownBestPosition[0]
self.topologyStats["position"] = self.bestParticle.position[:]
self.topologyStats["fitness"] = self.bestParticle.fitness
self.statted = True
def getBestParticle(self):
""" Return the best particle of the swarm
:rtype: the particle
"""
return self.bestParticle
def sort(self):
""" Sort the swarm """
if self.sorted: return
rev = (self.minimax == Consts.minimaxType["maximize"])
if self.sortType == Consts.sortType["fitness"]:
self.internalSwarm.sort(cmp=cmp_particle_fitness, reverse=rev)
self.sorted = True
def setSwarmSize(self, size):
""" Set the population size
:param size: the population size
"""
self.swarmSize = size
def setSortType(self,sort_type):
""" Sets the sort type
Example:
>>> swarm.setSortType(Consts.sortType["scaled"])
:param sort_type: the Sort Type
"""
self.sortType = sort_type
def setBestParticle(self,particle):
""" Set the best Particle
Example:
>>> topology.setBestParticle(bestParticle)
:param particle: the best particle to set
"""
self.bestParticle = particle
def create(self, **args):
""" Clone the example particle fo fill the swarm """
self.clear()
self.minimax = args["minimax"]
for i in xrange(self.swarmSize):
self.internalSwarm.append(self.oneSelfParticle.clone())
self.clear_flags()
def initialize(self):
""" Initialize all particles of swarm
this calls the initialize() of particles
PS: You can OVERRIDE this method for custom initialization.
"""
for particle in self.internalSwarm:
particle.initializePosition()
particle.initializeVelocity()
particle.evaluate()
particle.ownBestFitness = particle.fitness
self.bestParticle = self.internalSwarm[0]
self.clear_flags()
def evaluate(self,*args):
""" Evaluate all particles in swarm, calls the evaluate() method of particles
:param args: this param are passed to the evaluation function
"""
for particle in self.internalSwarm:
particle.evaluate(**args)
self.clear_flags()
def printStats(self):
""" Print statistics of the current population """
message = ""
#message = "[Swarm] - Max/Min/Avg bestFitness(Fitness) [%.2f(%.2f)/%.2f(%.2f)/%.2f(%.2f)]\n" % (self.stats["bestFitMax"], self.stats["fitMax"], self.stats["fitMin"], self.stats["bestFitMin"], self.stats["bestFitAvg"], self.stats["fitAvg"])
message+= "[Topology] - bestFitness/bestPosDim [%s/%s]" % (self.topologyStats["bestFitness"], self.topologyStats["bestPosDim"])
print message
return message
def clear(self):
""" Remove all particles from swarm """
del self.internalSwarm[:]
self.clear_flags()
def getStatistics(self):
""" Return a Statistics classes for statistics
:rtype: the subclasses of :class: `Statistics.Statistics` (topology,swarm) instances
"""
self.statistics()
self.clear_flags()
return (self.topologyStats, self.swarmStats)
