def main():
global ANSWERS
global DOMAIN
ANSWERS = mapping_benchmark.get_answers_by_qname(mapping_benchmark.SIM_READS_SAM_FILE)
#define min and max values, and arg for the numeric arguments we want to optimize:
DOMAIN = [
((1,300),'r'), #-r Score
((1,300),'q'), #-q mismatch Cost
((1,300),'a'), #-a gap existence cost
((1,300),'b'), #-b gap extension cost
#((1,300),'A'), #-A insertion existence cost, last says this is invalid??
#((1,300),'B'), #-B insertion extension cost, last says this is invalid??
#((),'c'), #-c gen affine gap costs? skip for now
((1,300),'x'), #-x drop
((1,300),'y'), #-y drop gaplees
((1,300),'z'), #-z drop final gapped
((1,300),'d'), #-d min score gapless
((1,300),'e'), #-e min alignment score
((1,300),'m'), #-m multipiclity
((30,100),'l'), #-l min length (less than 30 is way slow!)
#((),'r'), #-n count skip for now?
#((),'r'), #-t temperature?
#((),'r'), #-g gamma?
]
domain = [item[0] for item in DOMAIN]
final_scores = optimization.geneticoptimize(domain=domain, costf=cost,
popsize=300, step=3, mutprob=0.2, elite=0.2,
maxiter=10)
for (score,sol) in final_scores:
print score, gen_custom_args(sol)
def select_opt_method(self):
print self.flag
if self.flag==1:
return optimization.hillclimb(self.domain,self.fetcher.cost_func)
elif self.flag==2:
return optimization.annealingoptimize(self.domain,self.fetcher.cost_func)
elif self.flag==3:
return optimization.geneticoptimize(self.domain,self.fetcher.cost_func)
else:
return optimization.randomoptimize(self.domain,self.fetcher.cost_func)
def main():
domain = [(0, (len(dorms) * 2) - i - 1) for i in range(0, len(dorms) * 2)]
print("Random optimization:")
s = opt.randomoptimize(domain, dormcost)
printsolution(s)
print("\nSimulated annealing optimization:")
s = opt.annealingoptimize(domain, dormcost)
printsolution(s)
print("\nGenetic optimization:")
s = opt.geneticoptimize(domain, dormcost)
printsolution(s)
def Last_process(self):
self.Create_available_course_list()
if self.v.get() == 0:
solution_listt = optimization.hillclimb(self.domainn,
self.Cost_function)
self.write_listbox(solution_listt)
elif self.v.get() == 1:
solution_listt = optimization.annealingoptimize(
self.domainn, self.Cost_function)
self.write_listbox(solution_listt)
elif self.v.get() == 2:
solution_listt = optimization.geneticoptimize(
self.domainn, self.Cost_function)
self.write_listbox(solution_listt)
elif self.v.get() == 3:
solution_listt = optimization.randomoptimize(
self.domainn, self.Cost_function)
self.write_listbox(solution_listt)
def optimize(self):
self.consumers = [[con.strip()] for con in self.consumers.split(',')]
#append an empty list to each consumer to be filled by resources
for consumer in self.consumers:
consumer.append([]);
#What a solution looks like:
# [0,1,2,0,2,0,1,0,0,2,1,2,0,0,1,0,2,1,1,0,2,1,0...]
#where each list slot represents a particular resource
#from the resources list, and the value of a slot
#represents the consumer that resource is assigned to
#What the domain looks like
# [(0,2),(0,2),(0,2),(0,2),(0,2),(0,2),(0,2)...]
domain = [(0,len(self.consumers)-1)] * len(self.resources)
#Add up the total strength for each attribute possessed by a resource
total_attr_strengths = []
for a in range(0, len(self.resources[0]) - 1):
total_attr_strengths.append(sum([r[a+1] for r in self.resources]))
#Ideal attribute strength at each consumer.
#This is simple right now becuase we're just dividing total
#strengths by # of consumers.
self.optimal_attr_strengths = []
for a in range(0, len(total_attr_strengths)):
self.optimal_attr_strengths.append(total_attr_strengths[a]/len(self.consumers))
if self.algo == 'annealing':
cost, sol = optimization.annealingoptimize(domain, self.provisioning_cost)
elif self.algo == 'hillclimb':
cost, sol = optimization.hillclimb(domain, self.provisioning_cost)
elif self.algo == 'random':
cost, sol = optimization.randomoptimize(domain, self.provisioning_cost)
elif self.algo == 'genetic':
cost, sol = optimization.geneticoptimize(domain, self.provisioning_cost)
else:
raise ValueError('Invalid algorithm')
return cost, sol
draw = ImageDraw.Draw(img)
# 建立标识位置信息的字典
pos = dict([(people[i], (sol[i * 2], sol[i * 2 + 1]))
for i in range(0, len(people))])
for (a, b) in links:
draw.line((pos[a], pos[b]), fill=(255, 0, 0))
for n, p in pos.items():
draw.text(p, n, (0, 0, 0))
img.show()
domain = [(10, 370)] * (len(people) * 2) #设定题解范围
if __name__ == "__main__": #只有在执行当前模块时才会运行此函数
print(domain)
s = optimization.randomoptimize(domain, crosscount) # 随机搜索法，寻找最优题解
print(s)
drawnetwork(s) # 绘制关系网
s = optimization.hillclimb(domain, crosscount) # 爬山法，寻找最优题解
print(s)
drawnetwork(s) # 绘制关系网
s = optimization.annealingoptimize(domain, crosscount) # 模拟退火算法，寻找最优题解
print(s)
drawnetwork(s) # 绘制关系网
s = optimization.geneticoptimize(domain, crosscount) # 遗传算法，寻找最优题解
print(s)
drawnetwork(s) # 绘制关系网
print('\n<----Hill Climbing---->')
s = optimization.hillclimb(domain, optimization.schedulecost)
print(s)
cost = optimization.schedulecost(s)
print('Present cost:%f' % cost)
optimization.printschedule(s)
print('\n<----Simulated Annealing---->')
s = optimization.annealingoptimize(domain, optimization.schedulecost)
print(s)
cost = optimization.schedulecost(s)
print('Present cost:%f' % cost)
optimization.printschedule(s)
print('\n<----Genetic Algorithms---->')
s = optimization.geneticoptimize(domain, optimization.schedulecost)
print(s)
cost = optimization.schedulecost(s)
print('Present cost:%f' % cost)
optimization.printschedule(s)
# Student Dorm Problem
print('\n\n<----Student Dorm Optimization---->')
print('\n<----Random---->')
s = optimization.randomoptimize(dorm.domain, dorm.dormcost)
print('Present Cost:%d' % dorm.dormcost(s))
print('\n<----Genetic---->')
s = optimization.geneticoptimize(dorm.domain, dorm.dormcost)
print('Present Cost:%d' % dorm.dormcost(s))
dorm.printsolution(s)
# Loop over each student
for i in range(len(vec)):
x = int(vec[i])
dorm = dorms[slots[x]]
pref = prefs[i][1]
# First choice costs 0, second choice costs 1
if pref[0] == dorm:
cost += 0
elif pref[1] == dorm:
cost += 1
else:
cost += 3
# Not on the list costs 3
# Remove selected slot
del slots[x]
#print(cost)
return cost
print(domain)
s=optimization.randomoptimize(domain,dormcost)
printsolution(s)
s=optimization.hillclimb(domain,dormcost)
printsolution(s)
s=optimization.annealingoptimize(domain,dormcost)
printsolution(s)
s=optimization.geneticoptimize(domain,dormcost,elite=0.2)
printsolution(s)
# 遍历每一名学生
for i in range(len(vec)):
x = int(vec[i])
dorm = dorms[slots[x]]
pref = prefs[i][1]
# 首选成本值为0，次选成本值为1
if pref[0] == dorm: cost += 0
elif pref[1] == dorm: cost += 1
else: cost += 3
# 不在选择之列则成本值为3
# 删除选中的槽
del slots[x]
return cost
if __name__ == "__main__": #只有在执行当前模块时才会运行此函数
print(domain)
s = optimization.randomoptimize(domain, dormcost) # 随机搜索法，寻找最优题解
print(s)
printsolution(s) # 打印最优解代表的含义
s = optimization.hillclimb(domain, dormcost) # 爬山法，寻找最优题解
print(s)
printsolution(s) # 打印最优解代表的含义
s = optimization.annealingoptimize(domain, dormcost) # 模拟退火算法，寻找最优题解
print(s)
printsolution(s) # 打印最优解代表的含义
s = optimization.geneticoptimize(domain, dormcost) # 遗传算法，寻找最优题解
print(s)
printsolution(s) # 打印最优解代表的含义
def createschedule(people, dest, dep, ret):
sid = getkayaksession()
flights = {}
for p in people:
name, origin = p
# outbound flight
searchid = flightsearch(sid, origin, dest, dep)
flights[(origin, dest)] = flightsearchresults(sid, searchid)
# return flight
searchid = flightsearch(sid, dest, origin, ret)
flights[(origin, dest)] = flightsearchresults(sid, searchid)
return flights
if __name__ == '__main__':
import optimization
f = createschedule([('Nico', 'SFO'), ('Kerstin', 'SFO')],
'LGA',
dep='12/30/2008',
ret='1/2/2009')
domain = [(0, 30)] * len(f)
s = optimization.geneticoptimize(domain, optimization.costfunc(f, 'LGA'))
optimization.printschedule(s, 'LGA')
print('hillclimb algorithm -- schedulecost : ' )
print(optimization.schedulecost(s))
optimization.printschedule(s)
'''
'''
# annealing algorithm
s = optimization.annealingoptimize(domain, optimization.schedulecost)
print('annealing algorithm -- schedulecost : ' )
print(optimization.schedulecost(s))
optimization.printschedule(s)
'''
'''
print('---- genetic optimization algorithm ----')
s = optimization.geneticoptimize(domain, optimization.schedulecost)
print(s)
optimization.printschedule(s)
'''
s = optimization.randomoptimize(dorm.domain, dorm.dormcost)
print 'random optimize dormcost = ', dorm.dormcost(s)
dorm.printsolution(s)
s = optimization.geneticoptimize(dorm.domain, dorm.dormcost)
print 'genetic optimize dormcost = ', dorm.dormcost(s)
dorm.printsolution(s)
start = time.clock() s = op.randomoptimize(dorm.domain, costf=dorm.dormcost) print dorm.dormcost(s) dorm.printsolution(s) print 'running time', time.clock() - start print print 'hill climb:', start = time.clock() s = op.hillclimb(dorm.domain, costf=dorm.dormcost) print dorm.dormcost(s) dorm.printsolution(s) print 'running time', time.clock() - start print print 'annealing optimization:', start = time.clock() s = op.annealingoptimize(dorm.domain, costf=dorm.dormcost) print dorm.dormcost(s) dorm.printsolution(s) print 'running time', time.clock() - start print print 'genetic optimization:', start = time.clock() s = op.geneticoptimize(dorm.domain, costf=dorm.dormcost) print dorm.dormcost(s) dorm.printsolution(s) print 'running time', time.clock() - start print
# optimizationtest.py
import optimization
domain = [(0, 9)] * (len(optimization.people) * 2)
# hillclimb algorithm
'''
s = optimization.hillclimb(domain, optimization.schedulecost)
print('hillclimb algorithm -- schedulecost : ' )
print(optimization.schedulecost(s))
optimization.printschedule(s)
'''
'''
# annealing algorithm
s = optimization.annealingoptimize(domain, optimization.schedulecost)
print('annealing algorithm -- schedulecost : ' )
print(optimization.schedulecost(s))
optimization.printschedule(s)
'''
print('---- genetic optimization algorithm ----')
s = optimization.geneticoptimize(domain, optimization.schedulecost)
print(s)
optimization.printschedule(s)
def main(): s = optimization.geneticoptimize(domain, defendcost) printdefense(s)
print op.schedulecost(sol) op.printschedule(sol) print 'running time',time.clock()-start print print 'hill climb:', start = time.clock() sol = op.hillclimb(domain) print op.schedulecost(sol) op.printschedule(sol) print 'running time',time.clock()-start print print 'annealing optimization:', start = time.clock() sol = op.annealingoptimize(domain) print op.schedulecost(sol) op.printschedule(sol) print 'running time',time.clock()-start print print 'genetic optimization:', start = time.clock() sol = op.geneticoptimize(domain) print op.schedulecost(sol) op.printschedule(sol) print 'running time',time.clock()-start print
#coding:utf-8 import optimization import dorm import socialnetwork #s=[1,4,3,2,6,3,2,5,2,1,5,3] #optimization.printschedule(s) #print optimization.schedulecost(s) # optimization ²âÊÔ #domain = [(0,9)] * (len(optimization.people)*2) #s = optimization.randomoptimize(domain,optimization.schedulecost) #s = optimization.hillclimb(domain,optimization.schedulecost) #s = optimization.annealingoptimize(domain,optimization.schedulecost) #s = optimization.geneticoptimize(domain,optimization.schedulecost) #print optimization.schedulecost(s) #optimization.printschedule(s) #dorm ²âÊÔ #t = optimization.geneticoptimize(dorm.domain, dorm.dormcost) #dorm.printsolution(t) #socialnetwork ²âÊÔ sol = optimization.geneticoptimize(socialnetwork.domain, socialnetwork.crosscount) print socialnetwork.crosscount(sol) socialnetwork.drawnetwork(sol)
def Command_of_schedule_showing(self):
if len(
self.Course_code_shower.get(0, END)
) == 0: #If there is no course code in listbox,program will show an error.
self.Error_Message_Function(
"""Please First use the\nFetching Button""")
return
selections = map(int, self.Course_code_shower.curselection(
)) #Returning the list of indexes selected in the listbox
all_course_types = self.Course_code_shower.get(0, END)
selected_course_types = [all_course_types[i] for i in selections]
if len(
selected_course_types
) == 0: #If demanded course codes is not selected,program will show an error.
self.Error_Message_Function("""Please Select\nSome Course Code""")
return
#Returning the list of selected course codes.
self.List_of_all_avaiable_courses = list()
for course_types in selected_course_types: #Creating a list of courses from the course code dictionary with adding lists togetgher.
self.List_of_all_avaiable_courses.extend(
self.Dictionary_of_course_codes[course_types])
if len(
self.Course_number_taker.get()
) == 0: #If demanded course number is not provided,program will show an error.
self.Error_Message_Function("""Please give\nCourse number""")
return
try:
number_of_courses = int(self.Course_number_taker.get(
)) #Getting the demanded number of course from entry
except: #If it is not a number,program will comply.
self.Error_Message_Function(
"""Please give\nCourse number\nproperlyas integer""")
return
if number_of_courses > len(
self.List_of_all_avaiable_courses
): #If demeanded course number exceed avaiable course number,there will be an error.
self.Error_Message_Function(
"""These course codes\nhave insufficient course""")
return
self.domain = [(0, len(self.List_of_all_avaiable_courses) - i - 1)
for i in range(number_of_courses)]
#Domain
type_of_optimization = Optimization_Dictionary[self.Radio_Values.get()]
#Type of optimization come from the dictionary.
if type_of_optimization == "Hill Climbing":
solution = optimization.hillclimb(self.domain,
self.My_Cost_Function)
elif type_of_optimization == "Simulated Annealing":
solution = optimization.annealingoptimize(self.domain,
self.My_Cost_Function)
elif type_of_optimization == "Genetic Optimization":
if number_of_courses < 3: #For genetic optimization,we need at least three course demand
self.Error_Message_Function(
"""Genetic Optimization is\n valid for \nmore than 3 courses"""
)
return
solution = optimization.geneticoptimize(self.domain,
self.My_Cost_Function)
elif type_of_optimization == "Random Optimization":
solution = optimization.randomoptimize(self.domain,
self.My_Cost_Function)
Last_Showing = self.My_Representation_function(
solution
) #This returns a dictionary where keys are days and values will be list of tuple(time,course).
self.Last_Showing_Part.delete(0.0, END) #Deleting all the elements
for days in Days_of_Week: #List of days
self.Last_Showing_Part.insert(END, days + ":\n")
if days in Last_Showing: #If day inside of dictionary, inserting all tuples in the list(value of day).
for times, course in Last_Showing[days]:
self.Last_Showing_Part.insert(
END,
times + '(%s %s),' % (course.code, course.course_name))
else: #If day is not inside of the dictionary inserting no classes.
self.Last_Showing_Part.insert(END, 'NO CLASSES')
self.Last_Showing_Part.insert(END, '\n')
return total
def drawnetwork(sol):
# 建立image对象
img = Image.new('RGB', (400, 400), (255, 255, 255))
draw = ImageDraw.Draw(img)
# 建立标志位置信息的字典
pos = dict([(people[i], (sol[i * 2], sol[i * 2 + 1]))
for i in range(len(people))])
# 绘制连线
for (a, b) in links:
draw.line((pos[a], pos[b]), fill=(255, 0, 0))
# 绘制代表人的节点
for n, p in pos.items():
draw.text(p, n, (0, 0, 0))
img.show()
if __name__ == '__main__':
domain = [(10, 370)] * (len(people) * 2)
# sol = randomoptimize(domain, crosscount)
# sol = annealingoptimize(domain, crosscount, step=50, cool=0.99)
sol = geneticoptimize(domain, crosscount)
print crosscount(sol), sol
drawnetwork(sol)
x = int(vec[i])
dorm = dorms[slots[x]]
pref = prefs[i][1]
# first choice costs 0, second choice costs 1
if pref[0] == dorm:
cost += 0
elif pref[1] == dorm:
cost += 1
else:
# not on the list costs 3
cost += 3
# remove the selected slot
del slots[x]
return cost
if __name__ == "__main__":
# printsolution([0, 0, 0, 0, 0, 0, 0, 0, 0, 0])
s = optimization.randomoptimize(domain, dormcost)
print "Random searching", dormcost(s)
printsolution(s)
print '\n'
s = optimization.geneticoptimize(domain, dormcost)
print "Genetic algorithms", dormcost(s)
printsolution(s)
print '\n'
reload(optimization) domain = [(0,9)] * (len(optimization.people) * 2) s = optimization.randomoptimize(domain, optimization.schedulecost) optimization.schedulecost(s) optimization.printschedule(s) s = optimization.hillclimb(domain, optimization.schedulecost) optimization.schedulecost(s) optimization.printschedule(s) s = optimization.annealingoptimize(domain, optimization.schedulecost, T=10000, cool=0.95, step=3) optimization.schedulecost(s) optimization.printschedule(s) s = optimization.geneticoptimize(domain, optimization.schedulecost) optimization.printschedule(s) ####################### import dorm dorm.printsolution([0,0,0,0,0,0,0,0,0,0]) s = optimization.randomoptimize(dorm.domain, dorm.dormcost) dorm.dormcost(s) s = optimization.geneticoptimize(dorm.domain, dorm.dormcost) dorm.printsolution(s) ####################### import socialnetwork import optimization sol = optimization.randomoptimize(socialnetwork.domain, socialnetwork.crosscount)
def knn2(d, v):
return numpredict.weightedknn(d, v, k=3)
print(numpredict.crossvalidate(knn1, data))
print(numpredict.crossvalidate(knn2, data))
# Parameter tuning
import optimization
print('<----Parameter tuning---->')
data2 = numpredict.wineset2()
print(numpredict.crossvalidate(knn1, data2))
print(numpredict.crossvalidate(numpredict.weightedknn, data2))
print('<--scaling-->')
sdata = numpredict.rescale(data2, [10, 10, 0, 0.5])
print(numpredict.crossvalidate(knn1, sdata))
print('<--optimizing-->')
costf = numpredict.createcostfunction(numpredict.knnestimate, data2)
print(optimization.geneticoptimize(numpredict.weightdomain, costf, step=2))
# Probability Density
print('\n<----Probability Density---->')
data3 = numpredict.wineset3()
numpredict.cumulativegraph(data3, (99, 20), 120)
numpredict.probabilitygraph(data3, (99, 20), 120)
# eBay API
# remaining problems
# import ebaypredict
# laptops = ebaypredict.doSearch('laptop')
# print(laptops[:10])
[p.firstChild.data.split(' ')[1] for p in arrivals],
[parseprice(p.firstChild.data) for p in prices])
def createschedule(people, dest, dep, ret):
sid = getkayaksession()
flights = {}
for p in people:
name, origin = p
# outbound flight
searchid = flightsearch(sid, origin, dest, dep)
flights[(origin, dest)] = flightsearchresults(sid, searchid)
# return flight
searchid = flightsearch(sid, dest, origin, ret)
flights[(origin, dest)] = flightsearchresults(sid, searchid)
return flights
if __name__ == '__main__':
import optimization
f = createschedule([('Nico', 'SFO'), ('Kerstin', 'SFO')], 'LGA',
dep='12/30/2008', ret='1/2/2009')
domain = [(0, 30)] * len(f)
s = optimization.geneticoptimize(domain, optimization.costfunc(f, 'LGA'))
optimization.printschedule(s, 'LGA')
print 'hill climb:', start = time.clock() sol = op.hillclimb(so.domain,so.crosscount) print so.crosscount(sol) print 'running time',time.clock()-start print so.drawnetwork(sol) print 'annealing optimization:', start = time.clock() sol = op.annealingoptimize(so.domain,so.crosscount) print so.crosscount(sol) print 'running time',time.clock()-start print so.drawnetwork(sol) print 'genetic optimization:', start = time.clock() sol = op.geneticoptimize(so.domain,so.crosscount) print so.crosscount(sol) print 'running time',time.clock()-start print so.drawnetwork(sol)
def changedomain(domain, newdomain, sol):
return [
newdomain[i][0] + sol[i] / (domain[i][1] - domain[i][0]) *
(newdomain[i][1] - newdomain[i][0]) for i in range(len(domain))
]
def costf(sol):
newsol = changedomain(domain, newdomain, sol)
d = newsol[0] * 1e-3
epsr = newsol[1]
epsf = newsol[2]
w = 2 * math.pi * newsol[3] * 1e9
EPS = epsr - 1j * epsf
gamma = 1j * w / c * cmath.sqrt(EPS)
Zf = Z0 / cmath.sqrt(EPS) * cmath.tanh(gamma * d)
return abs((Zf - Z0) / (Zf + Z0))**2
sol = optimization.geneticoptimize(domain,
costf,
popsize=50,
step=1,
mutprob=0.2,
elite=0.2,
maxiter=100)
#sol=optimization.annealingoptimize(domain,costf,T=10000.0,cool=0.95,step=1)
#sol=optimization.randomhillclimb(domain,costf,N=100)
newsol = changedomain(domain, newdomain, sol)
print(newsol)
# op.printschedule(s)
# print(op.schedulecost(s))
domain = [(0, 9)] * (len(op.people) * 2)
# s = op.randomoptimize(domain, op.schedulecost)
# s = op.hillclimb(domain, op.schedulecost)
# s = op.annealingoptimize(domain, op.schedulecost)
# s = op.geneticoptimize(domain, op.schedulecost)
# print(domain,s)
# print(op.schedulecost(s))
# op.printschedule(s)
result = {'rand': [], 'hill': [], 'anne': [], 'gene': []}
for i in range(10):
s1 = op.randomoptimize(domain, op.schedulecost)
result['rand'].append(op.schedulecost(s1))
s2 = op.hillclimb(domain, op.schedulecost)
result['hill'].append(op.schedulecost(s2))
s3 = op.annealingoptimize(domain, op.schedulecost)
result['anne'].append(op.schedulecost(s3))
s4 = op.geneticoptimize(domain, op.schedulecost)
result['gene'].append(op.schedulecost(s4))
print(result)
print('rand', sum(result['rand']) / 10, min(result['rand']))
print('hill', sum(result['hill']) / 10, min(result['hill']))
print('anne', sum(result['anne']) / 10, min(result['anne']))
print('gene', sum(result['gene']) / 10, min(result['gene']))
def test_genetic(domain):
sys.stderr.write("testing genetic optimization...\n")
s = optimization.geneticoptimize(domain, optimization.schedulecost)
optimization.printschedule(s)
