def run_ga(target, control, water_stress, year):
print("Running Genetic Algorithm...")
elegidos_control = ga.ga(target, control)
print(
f"Selected control: {len(elegidos_control)} wavelength(s)\n{elegidos_control}"
)
rangos_control = ranges.rangos_clustering(target, elegidos_control,
"control", year, "ga")
print(f"Selected wavelength ranges (control set): {rangos_control}")
print("")
elegidos_water_stress = ga.ga(target, water_stress)
print(
f"Selected water stress: {len(elegidos_water_stress)} wavelength(s)\n{elegidos_water_stress}"
)
rangos_water_stress = ranges.rangos_clustering(target,
elegidos_water_stress, "ws",
year, "ga")
print(
f"Selected wavelength ranges (water stress set): {rangos_water_stress}"
)
control_long = data_preprocessing.wide_to_long(control.iloc[:, 1:])
graphics.ranges_graphics(target, control_long, rangos_control, "control",
year, "ga")
water_stress_long = data_preprocessing.wide_to_long(water_stress.iloc[:,
1:])
graphics.ranges_graphics(target, water_stress_long, rangos_water_stress,
"ws", year, "ga")
return
def main():
try:
args = parseArguments()
except:
args = {"-h": True}
if "-l" not in args:
args["-l"] = False
else:
args["-l"] = True
# handle argument
if "-h" in args:
printHelp()
return
if "-a" not in args:
greedy(args)
return
if args["-a"] == "aco":
aco(args)
elif args["-a"] == "ga":
ga(args)
elif args["-a"] == "greedy":
greedy(args)
elif args["-a"] == "mst":
mst(args)
else:
printHelp()
def update(name, dat, crash=None):
payload = {}
payload['an'] = __scriptname__
payload['av'] = __version__
payload['ec'] = name
payload['ea'] = 'zelka'
payload['ev'] = '1'
payload['dl'] = urllib.quote_plus(dat.encode('utf-8'))
ga().update(payload, crash)
def update(name, dat, crash=None):
payload = {}
payload['an'] = __scriptname__
payload['av'] = __version__
payload['ec'] = name
payload['ea'] = dat
payload['ev'] = '1'
payload['dl'] = None
ga().update(payload, crash)
def update(name, dat, crash=None):
payload = {}
payload["an"] = __scriptname__
payload["av"] = __version__
payload["ec"] = name
payload["ea"] = "zelka"
payload["ev"] = "1"
payload["dl"] = urllib.quote_plus(dat.encode("utf-8"))
ga().update(payload, crash)
def update(name, dat, crash=None):
payload = {}
payload['an'] = __scriptname__
payload['av'] = __version__
payload['ec'] = name
payload['ea'] = 'tv_service'
payload['ev'] = '1'
payload['dl'] = urllib.quote_plus(dat.encode('utf-8'))
ga('UA-79422131-11').update(payload, crash)
def update(name, dat, crash=None):
payload = {}
payload['an'] = __scriptname__
payload['av'] = __version__
payload['ec'] = name
payload['ea'] = dat
payload['ev'] = '1'
payload['dl'] = None
ga().update(payload, crash)
def update(name, dat, crash=None):
payload = {}
payload['an'] = __scriptname__
payload['av'] = __version__
payload['ec'] = name
payload['ea'] = 'sony_ctrl'
payload['ev'] = '1'
payload['dl'] = urllib.parse.quote_plus(dat)
ga().update(payload, crash)
def update(name, dat, crash=None):
payload = {}
payload['an'] = __scriptname__
payload['av'] = __version__
payload['ec'] = name
payload['ea'] = 'zamunda'
payload['ev'] = '1'
payload['dl'] = urllib.quote_plus(dat.encode('utf-8'))
ga().update(payload, crash)
def update(name, location, crash=None):
p = {}
p['an'] = addon.getAddonInfo('name')
p['av'] = addon.getAddonInfo('version')
p['ec'] = 'Addon actions'
p['ea'] = name
p['ev'] = '1'
p['ul'] = xbmc.getLanguage()
p['cd'] = location
ga('UA-79422131-7').update(p, crash)
def update(self, name, location, crash=None):
p = {}
p['an'] = self.addon_name
p['av'] = self.version
p['ec'] = 'Addon actions'
p['ea'] = name
p['ev'] = '1'
p['ul'] = xbmc.getLanguage()
p['cd'] = location
ga('UA-79422131-3').update(p, crash)
def update(name, location, crash=None):
p = {}
p['an'] = addon.getAddonInfo('name')
p['av'] = addon.getAddonInfo('version')
p['ec'] = 'Addon actions'
p['ea'] = name
p['ev'] = '1'
p['ul'] = xbmc.getLanguage()
p['cd'] = location
ga('UA-79422131-7').update(p, crash)
def update(self, name, location, crash=None):
p = {}
p['an'] = self.addon_name
p['av'] = self.version
p['ec'] = 'Addon actions'
p['ea'] = name
p['ev'] = '1'
p['ul'] = xbmc.getLanguage()
p['cd'] = location
ga('UA-79422131-3').update(p, crash)
def update(name, act_ev, dat, crash=None):
payload = {}
payload['ec'] = name
payload['ea'] = act_ev
payload['ev'] = '1'
payload['dl'] = urllib.quote_plus(dat.encode('utf-8'))
if run_from_xbmc == True:
payload['an'] = xbmcaddon.Addon().getAddonInfo('name')
payload['av'] = xbmcaddon.Addon().getAddonInfo('version')
ga().update(payload, crash)
else:
print payload
def update(name, act_ev, dat, crash=None):
payload = {}
payload['ec'] = name
payload['ea'] = act_ev
payload['ev'] = '1'
payload['dl'] = urllib.quote_plus(dat.encode('utf-8'))
if run_from_xbmc == True:
payload['an'] = xbmcaddon.Addon().getAddonInfo('name')
payload['av'] = xbmcaddon.Addon().getAddonInfo('version')
ga().update(payload, crash)
else:
print payload
def runOpt():
# X2 = GAP X3 = OVERLAP
# lb = [0.1, 0.05, 0.005,-0.03,0.5, 1.05,0.2]
lb = [0.03, 0.1, 0.03, 0.1, 0.005, -0.03]
ub = [0.20, 0.4, 0.20, 0.8, 0.030, 0.03]
GA=ga.ga(fg.flap_geometry2,lb,ub,nCPU=1,displayFlag=1)
GA.maxIter=5
GA=ga.ga(costFcn,lb,ub,0,8)
GA.Nelite=4
GA.maxIter=1000
GA.maxLagIter=50
GA.populationSize=30
GA.solve()
def __update__(action, location, crash=None):
try:
from ga import ga
p = {}
p['an'] = this.getAddonInfo('name').decode('utf-8')
p['av'] = this.getAddonInfo('version')
p['ec'] = 'Addon actions'
p['ea'] = action
p['ev'] = '1'
p['ul'] = xbmc.getLanguage()
p['cd'] = location
ga('UA-79422131-10').update(p, crash)
except Exception, er:
log(er)
def run_experiments():
for objective in objective_types:
for num_decision in num_decisions:
for num_objective in num_objectives:
problem = objective(num_decision, num_objective)
problem.find_min_max()
for type in types:
print objective.__name__ + ".%d.%d" % (
num_decision,
num_objective) + "." + type.__name__, "= [",
for repeat_ct in range(num_repeats):
random.seed(1131 + repeat_ct * 2 - 1)
print ga.ga(problem=problem, better=type), ",",
print "]"
def update(action, location, crash=None):
lu = settings.last_update
day = time.strftime("%d")
if lu == "" or lu != day:
settings.last_update
p = {}
p['an'] = get_addon_name()
p['av'] = get_addon_version()
p['ec'] = 'Addon actions'
p['ea'] = action
p['ev'] = '1'
p['ul'] = get_kodi_language()
p['cd'] = location
ga('UA-79422131-8').update(p, crash)
def update(name, location, crash=None):
lu = settings.last_update
d = time.strftime("%d")
if lu != d:
settings.last_update = d
p = {}
p['an'] = get_addon_name()
p['av'] = get_addon_version()
p['ec'] = 'Addon actions'
p['ea'] = name
p['ev'] = '1'
p['ul'] = xbmc.getLanguage()
p['cd'] = location
import ga
ga.ga('UA-79422131-12').update(p, crash)
def main():
# Load user parameters from file
config = ConfigParser()
config.read('params.ini')
# Choose function
function_name = config['PARAMS']['function_name']
if function_name == 'ackley':
fun = fun1
elif function_name == 'griewank':
fun = fun2
elif function_name == 'michalewicz':
fun = fun3
else:
raise Exception('Function not supported!')
# Do the algorithm
now = datetime.now()
x = ga(fun=fun,
num_of_population=int(config['PARAMS']['num_of_population']),
dimension=int(config['PARAMS']['dimension']),
num_of_iterations=int(config['PARAMS']['num_of_iterations']),
lower_bound=float(config['PARAMS']['lower_bound']),
upper_bound=float(config['PARAMS']['upper_bound']),
param_crossing=float(config['PARAMS']['param_crossing']),
param_mutation=float(config['PARAMS']['param_mutation']),
num_elitism=int(config['PARAMS']['elitism']))
# Results
print ('\n\nFunction results: \n')
print ('x = {0}\nf(x) = {1}\n'.format(x, fun(x)))
print ('Time: {0}'.format(datetime.now() - now))
def train(self):
print('GA run for tuning hyperparameters of OBJ. function...')
self.Theta, self.MinNegLnLik = ga(self.likelihood, self.nvars,
self.LowerTheta, self.UpperTheta)
print('GA run complete.')
self.U, self.mu, self.SigmaSqr = self.likelihood(self.Theta)[1:]
def run_first_mlp_ga():
'''
GA with a fitness function for determining the optimum number of hidden neurons and what inputs should be activated or deactivated.
Mutation rate determined by dividing 1 by the fitness function string length.
'''
pl.ion()
pl.show()
x = ga.ga(84, 'fF.fitness_mlp1', 100, 50, 0.01, 4, True)
x.runGA()
def objectives(self):
model = dtlz(10, 2, 1)
print "Calling GA with decs", self.decs
calculated_objs = ga(model, self.base_min, self.base_max, self.basepop, self.decs[2], self.decs[3], self.decs[0], self.decs[1]) #args = (candidates, generation, mutation, crossover)
#print "new objs", calculated_objs
return calculated_objs
def objectives(self):
model = dtlz(10, 2, 1)
print "Calling GA with decs", self.decs
calculated_objs = ga(
model, self.base_min, self.base_max, self.basepop, self.decs[2],
self.decs[3], self.decs[0], self.decs[1]
) #args = (candidates, generation, mutation, crossover)
#print "new objs", calculated_objs
return calculated_objs
def __update__(action, location, crash=None):
try:
lu = settings.last_update
day = time.strftime("%d")
if lu != day:
settings.last_update = day
from ga import ga
p = {}
p['an'] = this.getAddonInfo('name').decode('utf-8')
p['av'] = this.getAddonInfo('version')
p['ec'] = 'Addon actions'
p['ea'] = action
p['ev'] = '1'
p['ul'] = xbmc.getLanguage()
p['cd'] = location
ga('UA-79422131-10').update(p, crash)
except Exception, er:
log(er)
def run_second_mlp_ga():
'''
GA with a fitness function for determining the optimum learning rate and number of iterations.
Mutation rate determined by dividing 1 by the fitness function string length.
'''
pl.ion()
pl.show()
x = ga.ga(25, 'fF.fitness_mlp2', 50, 50, 0.05, 4, True)
x.runGA()
def __update__(action, location, crash=None):
try:
import time
lu = settings.last_update
day = time.strftime("%d")
if lu != day:
settings.last_update = day
from ga import ga
p = {}
p['an'] = get_addon_name()
p['av'] = get_addon_version()
p['ec'] = 'Addon actions'
p['ea'] = action
p['ev'] = '1'
p['ul'] = get_system_language()
p['cd'] = location
ga('UA-79422131-10').update(p, crash)
except Exception, er:
log(er)
def get_objectives(self, solution):
self.init_hve_cache, hvol = ga(self.problem,
pop_size=solution[3],
gens=solution[4],
select_type=solution[2],
mutation_rate=solution[0],
crossover_type=solution[1],
init_hve=self.init_hve_cache)
self.update_min_max(hvol)
return hvol
def stage_11():
lb = array([0.05,0.001,0.001,0.001,-0.4 ,-0.4 ,-0.4])
ub = array([0.3 ,0.4 ,0.4 ,0.4 , 0.1, 0.1, 0.1])
design = Design()
costFcn = design.objective
GA = ga.ga(costFcn,lb,ub)
GA.setCPU(1)
GA.displayFlag = 1
GA.solve()
print 'f=%.4f x1=%.4f x2=%.4f, Neval=%d' %(GA.fOpt,GA.xOpt[0],GA.xOpt[1],GA.funcCount)
def yplacement(drawing_object_dict, connection_list, inst_col_dict):
""" Place the instanciations of the current module in the y-axis.
At the minute, this is pure combinatorial. Will take ages for big ccts.
"""
global drawing_object_dict_ref
global inst_col_dict_ref
global connection_list_ref
global drawing_object_name_list
global possible_crossovers
# Determine the number of objects to place so that we can set up the GA
num_drawing_objects = len(drawing_object_dict)
# Set up references to the drawing object dictionary
drawing_object_dict_ref = drawing_object_dict
inst_col_dict_ref = inst_col_dict
connection_list_ref = connection_list
# A sorted list of block names is needed so that we can assign genes
# by position to y-axis settings of the blocks.
drawing_object_name_list = drawing_object_dict.keys()
drawing_object_name_list.sort()
# Determine which nets can possible cross each other
connection_point_coord_list = find_pin_coords(connection_list,
drawing_object_dict,
inst_col_dict)
possible_crossovers = find_possible_crossovers(
connection_list, connection_point_coord_list)
# Configure the Genetic Algorithm
placement_ga = ga.ga(
fitness_function=layout_fitness_function,
num_genes=num_drawing_objects,
num_generations=20,
population_size=120,
num_crossovers=2,
num_elite=20,
num_parents=28,
mutation_rate=0.01,
max_range=16,
mutation_max_deviation=2,
go_low=True,
)
# Run the GA, then choose the fittest member of the DGA population as
# the final layout values
print "-" * 80
best_layout = placement_ga.evolve()
#
return
def yplacement( drawing_object_dict, connection_list, inst_col_dict ):
""" Place the instanciations of the current module in the y-axis.
At the minute, this is pure combinatorial. Will take ages for big ccts.
"""
global drawing_object_dict_ref
global inst_col_dict_ref
global connection_list_ref
global drawing_object_name_list
global possible_crossovers
# Determine the number of objects to place so that we can set up the GA
num_drawing_objects = len( drawing_object_dict )
# Set up references to the drawing object dictionary
drawing_object_dict_ref = drawing_object_dict
inst_col_dict_ref = inst_col_dict
connection_list_ref = connection_list
# A sorted list of block names is needed so that we can assign genes
# by position to y-axis settings of the blocks.
drawing_object_name_list = drawing_object_dict.keys()
drawing_object_name_list.sort()
# Determine which nets can possible cross each other
connection_point_coord_list = find_pin_coords(connection_list, drawing_object_dict, inst_col_dict)
possible_crossovers = find_possible_crossovers(connection_list, connection_point_coord_list)
# Configure the Genetic Algorithm
placement_ga = ga.ga(
fitness_function = layout_fitness_function,
num_genes = num_drawing_objects,
num_generations = 20,
population_size = 120,
num_crossovers = 2,
num_elite = 20,
num_parents = 28,
mutation_rate = 0.01,
max_range = 16,
mutation_max_deviation = 2,
go_low = True,
)
# Run the GA, then choose the fittest member of the DGA population as
# the final layout values
print "-"*80
best_layout = placement_ga.evolve()
#
return
def Ok_command(self):
self.initCitys()
self.get_CityOrder()
self.ga = ga(StartCity=self.CityOrder[0],
gene=self.CityOrder,
PCross=0.8,
PMutation=0.02,
LifeCount=100,
GeneLength=len(self.CityOrder),
MatchFun=self.matchFun())
n = 100
while n > 1:
self.ga.next()
distance = self.distance(self.ga.best.gene)
n -= 1
self.ltt.insert(tk.INSERT, '最佳路径为：\n')
for i in range(len(self.ga.best.gene) - 1):
self.ltt.insert(
tk.INSERT, self.citys[self.ga.best.gene[i]][0] + '-------->' +
self.citys[self.ga.best.gene[i + 1]][0] + ' ' + str(
geodesic(
(self.citys[self.ga.best.gene[i]][2],
self.citys[self.ga.best.gene[i]][1]),
(self.citys[self.ga.best.gene[i + 1]][2],
self.citys[self.ga.best.gene[i + 1]][1]))) + '\n')
self.ltt.insert(
tk.INSERT, self.citys[self.ga.best.gene[-1]][0] + '-------->' +
self.citys[self.ga.best.gene[0]][0] + ' ' + str(
geodesic((self.citys[self.ga.best.gene[-1]][2],
self.citys[self.ga.best.gene[-1]][1]),
(self.citys[self.ga.best.gene[0]][2],
self.citys[self.ga.best.gene[0]][1]))) + '\n')
self.ltt.insert(tk.INSERT,
'路程共计为：\n' + str(self.distance(self.CityOrder)))
self.line()
def __init__(self,
population,
function,
inputs,
hidden,
outputs,
display=True,
mutation_rate=0.01,
survival_rate=0.3,
mutation_power=0.3,
species=1,
save_to_file=False,
best_mode=False):
self.population = population
self.func = function
self.fitness_sums = []
self.species = species
self.best_mode = best_mode
self.save_to_file = save_to_file
if self.save_to_file and FileHandler.read('generation.pickle'):
self.nets = FileHandler.read('generation.pickle')
if (self.best_mode and FileHandler.read('best.pickle')):
self.best = FileHandler.read('best.pickle')
self.nets = [[self.best]]
elif not self.save_to_file or not FileHandler.read(
'generation.pickle'):
if best_mode:
print("Could not find a file, acting ordirnally")
self.nets = [[
ga(inputs, hidden, outputs, mutation_rate, mutation_power)
for i in range(int(self.population / species))
] for i in range(species)]
self.best = None
self.p = None
self.survival_rate = survival_rate
self.display = display
best_sol, best_cost, data = tb.tb(
n,
adj_mat,
tb_size=20, # tabu solutions in tb_list
max_tnm=100, # how many candidates picked in tournament selection
mut_md=
mut_md, # [get_sol, get delta], method of mutation, e.g. swap, 2-opt
term_count=200 # terminate threshold if best_cost nor change
)
elif method == 'ga':
method_name = 'Genetic Algorithm'
best_sol, best_cost, data = ga.ga(
n,
adj_mat,
n_pop=200,
r_cross=0.5,
r_mut=0.8,
selection_md='tnm', # 'rw' / 'tnm' / 'elt'
max_tnm=3,
term_count=200)
elif method == 'sa':
method_name = 'Simulated Annealing'
best_sol, best_cost, data = sa.sa(
n,
adj_mat,
tb_size=0, # tabu solutions in tb_list
max_tnm=20, # how many candidates picked in tournament selection
mut_md=
mut_md, # [get_sol, get delta], method of mutation, e.g. swap, 2-opt
term_count_1=25, # inner loop termination flag
term_count_2=25, # outer loop termination flag
def step(self):
for el in self.elevator:
for key in el.timer:
if el.timer[key] >= 0:
el.timer[key] -= 1
# elevator ID counter
el_id = 0
# loop through every elevator in self.elevator
for el in self.elevator:
el.setHeader()
# loop through every key in el.timer dictionary
for key in el.timer:
# timer equal to 0 it means some action have to occur
if el.timer[key] == 0:
el.timer[key] = -1
# switch to find which is the key equal to 0
# if the key is 'moving'
if key == 'moving':
# then increment or decremtn the floor with regards to
# the direction (up or down respectively)
curr_floor = el.current_floor
if el.destination_floor > el.current_floor:
el.current_floor += 1
elif el.destination_floor < el.current_floor:
el.current_floor -= 1
else:
# if the direction is neither up nor down
# then do nothing
pass
self.signals["setElevatorFloor"].emit(
curr_floor, el.current_floor, el_id)
# if there is one or more passenger in the elevator who need to get off
# or if there are up or down calls at the current floor
if len(el.passengersGettingOff()) > 0 or len(
self.passengersGettingOn(el_id)) > 0:
# then stop (that is decelerate, stop and open doors)
el.moveToStop()
else:
# else continue moving upward or downward
el.move()
elif key == 'move_to_stop':
# if there is one or more passenger in the elevator who needs to get off
if len(el.passengersGettingOff()) > 0:
el.unload()
# else if there is one or more passenger who needs to get on
elif len(self.passengersGettingOn(el_id)):
el.load()
elif key == 'stop_to_move':
# azione di movimento
el.move()
elif key == 'loading':
# p = self.floor_queue[el.current_floor].dequeue()
# el.passenger.append(p)
# el.timer['stop_to_move'] = H
self.getOn(el_id)
el.stopToMove()
if self.assignment[el.current_floor] == el_id:
self.assignment[el.current_floor] = -1
if self.assignment[int(el.current_floor +
len(self.assignment) / 2) -
1] == el_id:
self.assignment[int(el.current_floor +
len(self.assignment) / 2) -
1] = -1
self.updateElevatorsDestinationFloor()
elif key == 'unloading':
# p = el.passenger.dequeue()
# self.floor_queue[el.current_floor].enqueue(p)
# el.timer['move_to_stop'] = L
el.getOff()
if len(self.passengersGettingOn(el_id)) > 0:
el.load()
self.updateElevatorsDestinationFloor()
elif el.current_floor != el.destination_floor:
el.stopToMove()
else:
raise KeyError(
"Unknown elevator timer key in step function")
el_id += 1
#Se abbiamo nuove chiamate:
if self.new_calls:
if DEBUG:
print("GA will parse new calls")
# Passive Time
pt = SETTINGS["elevator"]["timing"]["loading"] + SETTINGS[
"elevator"]["timing"]["move_to_stop"] + SETTINGS["elevator"][
"timing"]["stop_to_move"]
# Inter floor trip time
it = SETTINGS["elevator"]["timing"]["moving"]
# Hall call UP/DOWN
hcu = []
hcd = []
call_flag = False
for i in range(len(self.floor_queue)):
queue = self.floor_queue[i]
if i < len(self.floor_queue) - 1:
hcu.append(0)
if i > 0:
hcd.append(0)
for p in queue:
call_flag = True
if p.destination_floor > p.origin_floor:
hcu[i] = 1
elif p.destination_floor < p.origin_floor:
hcd[i - 1] = 1
else:
raise Exception(
"Passenger destination_floor == origin_floor")
cf = []
cdf = []
for el in self.elevator:
cf.append(el.current_floor)
cdf.append(el.destination_floor)
try:
'''
print(
"\n****************************************",
"\nhcu=" + str(hcu),
"\nhcd=" + str(hcd),
"\ncf=" + str(cf),
"\ncdf=" + str(cdf),
"\n****************************************"
)
input()
'''
if call_flag:
self.assignment = ga.ga(self.nf, self.nc, pt, it,
list(hcu), list(hcd), cf,
cdf).computeSolution()
except Exception as e:
print("GA error DEBUG", "hcu=" + str(list(hcu)),
"hcd=" + str(list(hcd)), "cf=" + str(cf),
"cdf=" + str(cdf))
raise e
self.new_calls = False
self.updateElevatorsDestinationFloor()
for el_id in range(len(self.elevator)):
el = self.elevator[el_id]
if el.isIdle():
if el.destination_floor != el.current_floor:
el.stopToMove()
else:
if len(self.passengersGettingOn(el_id)) > 0:
el.load()
else:
for i in range(len(self.floor_queue)):
queue = self.floor_queue[i]
upgoings = 0
downgoings = 0
for p in queue:
if p.destination_floor > p.origin_floor:
upgoings += 1
else:
downgoings += 1
if i == 0:
up_assign = self.assignment[i]
if up_assign == -1 and upgoings > 0:
self.new_calls = True
break
elif 0 < i < len(self.floor_queue) - 1:
down_assign = self.assignment[
i + int(len(self.assignment) / 2) - 1]
if down_assign == -1 and downgoings > 0:
self.new_calls = True
break
up_assign = self.assignment[i]
if up_assign == -1 and upgoings > 0:
self.new_calls = True
break
elif i == len(self.floor_queue) - 1:
down_assign = self.assignment[
i + int(len(self.assignment) / 2) - 1]
if down_assign == -1 and downgoings > 0:
self.new_calls = True
break
self.__t = m_start(self.__evt)
if __name__ == '__main__':
if __addon__.getSetting('firstrun') == 'true':
__addon__.openSettings()
__addon__.setSetting('firstrun', 'false')
if __addon__.getSetting('dbg') == 'true':
fname = 'dbg.png'
else:
fname = None
payload = {}
payload['an'] = __scriptname__
payload['av'] = __version__
payload['ec'] = 'led_start'
payload['ea'] = 'led_start'
payload['ev'] = '1'
ga().update(payload, None)
monitor = MyMonitor()
while True:
# Sleep/wait for abort for 1 seconds
if monitor.waitForAbort(1):
# Abort was requested while waiting. We should exit
break
del monitor
def next_infill(self):
print('GA run for optimizing EI...')
x, EI = ga(self.exp_imp, self.nvars, np.zeros((self.nvars, )), np.ones((self.nvars, )))
print('GA run complete.')
return x, EI
# Code from Chapter 10 of Machine Learning: An Algorithmic Perspective (2nd Edition) # by Stephen Marsland (http://stephenmonika.net) # You are free to use, change, or redistribute the code in any way you wish for # non-commercial purposes, but please maintain the name of the original author. # This code comes with no warranty of any kind. # Stephen Marsland, 2008, 2014 # A runner for the Genetic Algorithm import ga import pylab as pl pl.ion() pl.show() plotfig = pl.figure() ga = ga.ga(30, "fF.fourpeaks", 301, 100, -1, "un", 4, True) ga.runGA(plotfig) pl.pause(0) # pl.show()
for MmutationProb in range(2):
if MmutationProb == 0:
mutationProb = -1.0
else:
mutationProb = MmutationProb / 10.0
success = []
fail = []
for testI in range(20):
ga1 = ga.ga(
stringLength * (MstringLength + 1),
fitnessFunction, nEpochs * (MnEpochs + 1),
populationSize * (MpopulationSize + 1),
mutationProb, crossover,
nElite * (MnElite + 1), tournament)
Fitness = ga1.runGA()
dict = {}
#dict["iteration"] =
f = list(filter(lambda x: x > 100, Fitness))
localFail = list(
filter(lambda x: x <= 100, Fitness))
success.extend(f)
fail.extend(localFail)
#print("\ni = ", i)
#print("Fitness", Fitness)
params = []
sses_ga = []
for ite in range(nsim):
ts = np.arange(-0, len(dates) - index_2020_10_13 - 0 - 1, dt)
varbound = np.array([[0, 0.1], [0, 0.01], [4, 6], [1, 4]])
algorithm_param = {'max_num_iteration':niter,\
'population_size': npop,\
'mutation_probability':0.1,\
'elit_ratio': 0.01,\
'crossover_probability': 0.5,\
'parents_portion': 0.3,\
'crossover_type':'uniform',\
'max_iteration_without_improv':None}
model = ga(function=model_evaluation_param_est,
dimension=4,
variable_type='real',
variable_boundaries=varbound,
algorithm_parameters=algorithm_param)
param_opt = model.run()
param = model.output_dict['variable']
sses_ga.append(model.output_dict['function'])
params.append(param)
# @jit(nopython=True, parallel=True)
# def main(nsim):
# params = np.zeros((nsim,4),dtype=np.float64)
# sses_ga=np.zeros(nsim,dtype=np.float64)
# ts = np.arange(-0, len(dates)-index_2020_10_13-0-1, dt)
# for ite in prange(nsim):
# varbound=np.array([[0,0.1],[0,0.01],[4,6],[1,4]])
# algorithm_param = {'max_num_iteration':niter,\
# Code from Chapter 12 of Machine Learning: An Algorithmic Perspective # by Stephen Marsland (http://seat.massey.ac.nz/personal/s.r.marsland/MLBook.html) # You are free to use, change, or redistribute the code in any way you wish for # non-commercial purposes, but please maintain the name of the original author. # This code comes with no warranty of any kind. # Stephen Marsland, 2008 # A runner for the Genetic Algorithm import ga ga = ga.ga(20,'fF.knapsack',101,100,-1,'sp',4,True) ga.runGA()
import ga import numpy as np import pylab as pl popSize = 5 maxHiddenNodes= 32 # (43 plus one datapoints) stringLength = 45*maxHiddenNodes pop = np.random.rand(popSize, stringLength) pop = np.where(pop<0.5,1,1) pl.ion() pl.show() plotfig = pl.figure() ga = ga.ga(stringLength,'fF.myGA',1001,popSize,0.1,'none',0,False) ga.runGA(plotfig) pl.pause(0) pl.show()
# Code from Chapter 10 of Machine Learning: An Algorithmic Perspective (2nd Edition) # by Stephen Marsland (http://stephenmonika.net) # You are free to use, change, or redistribute the code in any way you wish for # non-commercial purposes, but please maintain the name of the original author. # This code comes with no warranty of any kind. # Stephen Marsland, 2008, 2014 # A runner for the Genetic Algorithm import pylab as pl import ga pl.ion() pl.show() plotfig = pl.figure() ga = ga.ga(30, 'fF.fourpeaks', 301, 100, -1, 'un', 4, True) ga.runGA(plotfig) pl.pause(0) # pl.show()
punishments += two_exams_at_once * STUDENT_TAKING_TWO_EXAMS_AT_ONCE
punishments += exam_after_another * STUDENT_TAKING_EXAMS_ONE_AFTER_ANOTHER
punishments += more_than_two_exams_at_one_day * STUDENT_TAKING_MORE_THAN_TWO_EXAMS_IN_ONE_DAY
punishments += two_exams_at_one_day * STUDENT_TAKING_TWO_EXAMS_IN_ONE_DAY
if inverse:
return 1.0 / (1.0 + punishments),
else:
is_valid = two_exams_at_once == 0
return punishments, is_valid, {
"two exams at once": two_exams_at_once,
"exam after another": exam_after_another,
"more than two exams at one day": more_than_two_exams_at_one_day,
"two exams at one day": two_exams_at_one_day
}
# Launch this only if this script is launched directly (don't launch this if `from main import *` is invoked)
if __name__ == '__main__':
ga(pop_count=POPULATION_COUNT,
generations_number=GENERATIONS_NUMBER,
cx_prob=CROSSOVER_PROBABILITY,
mut_prob=MUTATION_PROBABILITY,
mut_change_exam_prob=MUTATION_CHANGE_EXAM_PROBABILITY,
evaluate_func=evaluate,
available_timeslots=available_timeslots,
exams=exams,
timeslot_to_day=timeslot_to_day,
timeslot_to_dayslot=timeslot_to_dayslot,
print_best=True)
def train(self):
print('GA run for tuning hyperparameters of OBJ. function...')
self.Theta, self.MinNegLnLik = ga(self.likelihood, self.nvars, self.LowerTheta, self.UpperTheta)
print('GA run complete.')
self.U, self.mu, self.SigmaSqr = self.likelihood(self.Theta)[1:]
import ga ## initialise the GA oga = ga.ga() oga.seedGA(2, 10, 8, 4) pops = oga.getPopulation() print(pops[0]["first_weights"].renderT()) print(pops[1]["first_weights"].renderT()) print(oga.mutatePopulation(pops, 1, 1)) pops = oga.getPopulation() print(pops[0]["first_weights"].renderT()) print(pops[1]["first_weights"].renderT())
from pso import PSO
import ga
from test_func import sphere as obj_func
print('------------')
print('starting PSO...')
# my_pso = PSO(func=obj_func, pop=30, dim=5, max_iter=100)
my_pso = PSO(func=obj_func, dim=3)
fitness = my_pso.fit()
print('best_x is ', my_pso.gbest_x)
print('best_y is ', my_pso.gbest_y)
my_pso.plot_history()
print('-------------')
print('starting GA...')
# general_best,func_general_best=ga.ga(func=demo_func2, pop=50, iter_max=200, lb=[-1, -10, -5], ub=[2, 10, 2],precision=[1e-7, 1e-7, 1e-7],Pm=0.001)
# print(general_best,func_general_best)
general_best, func_general_best, FitV_history = ga.ga(func=obj_func,
lb=[-1, -10, -5],
ub=[2, 10, 2])
print('best_x:', general_best)
print('best_y:', func_general_best)
ga.plot_FitV(FitV_history)
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE # ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE # LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR # CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF # SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS # INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN # CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) # ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE # POSSIBILITY OF SUCH DAMAGE. import ga, Matrix, Ann, copy, random ## initialise the GA oga = ga.ga() oga.seedGA(5, 20, 5, 4) for p in oga.getPopulation(): print(p["second_thresholds"]) print(p["second_thresholds"].renderT()) pop = oga.getPopulation() newPop = [] for n in range(5): newPop.append(copy.deepcopy(pop[0])) currVal = newPop[n]["second_thresholds"].getElem(1, 0) currVal += (0.1 - random.random()*0.2) newPop[n]["second_thresholds"].setElem(1, 0, currVal)
import numpy as np
import ga as ga
import time
def costFcn(x,params):
import numpy as np
for i in range(10000):
g=np.array([0,0])
f=100*(x[1]-x[0]**2)**2+(1-x[0])**2
g[0]=2*x[0]-x[1]
g[1]=1-x[0]
return (f,g)
lb=np.array([-2.0,-5.0])
ub=np.array([3.0,5.0])
x=np.array([2.0,1.0])
f,g=costFcn(x,0.0)
GA=ga.ga(costFcn,lb,ub)
# denorm test
XNorm=np.random.rand(100,2)#np.array([[0.5,0.4],[1.0,1.0],[0.0,0.0]])
X=GA._denorm(XNorm)
time1=time.time()
F,G=GA._systemEval(X)
print "Single core time = ",time.time()-time1, "s"
#print " input X values = ",X
#print "output F values = ",F
#print "output G values = ",G
time2=time.time()
F,G=GA._systemEval_pp(X)
print "Multi core time = ",time.time()-time2, "s"
#print " input X values = ",X
