def __init__(prob):
prob.name = "Osyczka2"
lows = [0, 0, 1, 1, 0, 0]
ups = [10, 10, 5, 5, 6, 10]
names = ["x1", "x2", "x3", "x4", "x5", "x6"]
prob.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(6)]
prob.objectives = [jmoo_objective("f1", True), jmoo_objective("f2", True)]
def __init__(prob):
super(zdt1, prob).__init__()
prob.name = "ZDT1"
names = ["x" + str(i+1) for i in range(30)]
prob.decisions = [jmoo_decision(names[i], 0, 1) for i in range(len(names))]
prob.objectives = [jmoo_objective("f1", True), jmoo_objective("f2", True)]
def __init__(prob):
prob.name = "ZDT4"
names = ["x" + str(i+1) for i in range(10)]
prob.decisions = [jmoo_decision(names[i], -5, 5) for i in range(len(names))]
prob.decisions[0].low = 0
prob.decisions[0].up = 1
prob.objectives = [jmoo_objective("f1", True), jmoo_objective("f2", True)]
def __init__(self, requirements, releases, clients, density, budget):
self.name = "NRP_" + str(requirements) + "_" + str(releases) + "_" + str(clients) + "_" +str(density) \
+ "_" + str(budget)
names = ["x" + str(i + 1) for i in range(requirements)] # |x_i + y_i|
lows = [-1 for i in xrange(requirements)]
ups = [(releases - 1) for _ in xrange(requirements)]
self.decisions = [
jmoo_decision(names[i], lows[i], ups[i])
for i in range(requirements)
]
self.objectives = [
jmoo_objective("f1", False),
jmoo_objective("f2", True),
jmoo_objective("f3", False)
]
self.trequirements = requirements
self.treleases = releases
self.tclients = clients
self.tdensity = density
self.tbudget = budget
self.requirement = None
self.client = None
self.release = None
self.precedence = []
self.generate_data()
def __init__(prob):
prob.name = "POM3MIN"
names = ["Culture", "Criticality", "Criticality Modifier", "Initial Known", "Inter-Dependency", "Dynamism", "Size", "Plan", "Team Size"]
LOWS = [0.80, 1.22, 2, 0.60, 0, 1, 0, 0, 1]
UPS = [0.90, 1.62, 6, 0.62, 2, 3, 1, 1, 3]
prob.decisions = [jmoo_decision(names[i], LOWS[i], UPS[i]) for i in range(len(names))]
prob.objectives = [jmoo_objective("Cost", True), jmoo_objective("Score", False), jmoo_objective("Completion", False), jmoo_objective("Idle", True)]
def __init__(prob):
prob.name = "ScaleTestDissimilar"
prob.decisions = [jmoo_decision("x1", 0, 1), jmoo_decision("x2", 0, 1)]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob):
prob.name = "Golinski"
names = ["x1", "x2", "x3", "x4", "x5", "x6", "x7"]
LOWS = [2.6, 0.7, 17.0, 7.3, 7.3, 2.9, 5.0]
UPS = [3.6, 0.8, 28.0, 8.3, 8.3, 3.9, 5.5]
prob.decisions = [jmoo_decision(names[i], LOWS[i], UPS[i]) for i in range(7)]
prob.objectives = [jmoo_objective("f1", True), jmoo_objective("f2", True)]
def __init__(prob, n):
prob.evals = 0
prob.name = "Fonseca"
names = ["x1", "x2", "x3", "x4", "x5", "x6", "x7", "x8"]
n = max(min(8, n), 1) #constrain n to its range
prob.decisions = [jmoo_decision("x" + str(i+1), -2, 2) for i in range(n)]
prob.objectives = [jmoo_objective("f1", True), jmoo_objective("f2", True)]
def __init__(prob):
prob.name = "POM3Asanscomp"
names = ["Culture", "Criticality", "Criticality Modifier", "Initial Known", "Inter-Dependency", "Dynamism", "Size", "Plan", "Team Size"]
LOWS = [0.1, 0.82, 2, 0.40, 1, 1, 0, 0, 1]
UPS = [0.9, 1.20, 10, 0.70, 100, 50, 4, 5, 44]
prob.decisions = [jmoo_decision(names[i], LOWS[i], UPS[i]) for i in range(len(names))]
prob.objectives = [jmoo_objective("Cost", True, 0), jmoo_objective("Score", False, 0, 1), jmoo_objective("Idle", True, 0, 1)]
def __init__(prob):
prob.name = "POM3D"
names = ["Culture", "Criticality", "Criticality Modifier", "Initial Known", "Inter-Dependency", "Dynamism", "Size", "Plan", "Team Size"]
LOWS = [0.10, 0.82, 2, 0.60, 80, 1, 0, 0, 10]
UPS = [0.20, 1.26, 8, 0.95, 100, 10, 2, 5, 20]
prob.decisions = [jmoo_decision(names[i], LOWS[i], UPS[i]) for i in range(len(names))]
prob.objectives = [jmoo_objective("Cost", True, 0), jmoo_objective("Score", False, 0, 1), jmoo_objective("Completion", False, 0, 1), jmoo_objective("Idle", True, 0, 1)]
def __init__(prob):
prob.name = "BNH"
prob.decisions = [jmoo_decision("x1", 0, 5), jmoo_decision("x2", 0, 3)]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob):
prob.name = "Schaffer"
prob.decisions = [jmoo_decision("x1", -10, 10)]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob):
prob.name = "BiasTest"
prob.decisions = [jmoo_decision("x1", 0, 1), jmoo_decision("x2", 0, 1)]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob):
prob.name = "A001_Test"
prob.A = 0.01
prob.decisions = [jmoo_decision("x1", 0, 1)]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob, n):
prob.name = "JoeTest1"
names = ["x1", "x2", "x3"]
prob.decisions = [jmoo_decision(names[i], -2, 2) for i in range(n)]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob):
prob.name = "Srinivas"
prob.decisions = [
jmoo_decision("x" + str(i + 1), -20, 20) for i in range(2)
]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob):
prob.name = "POM3MIN"
names = ["Culture", "Criticality", "Criticality Modifier", "Initial Known", "Inter-Dependency", "Dynamism",
"Size", "Plan", "Team Size"]
LOWS = [0.80, 1.22, 2, 0.60, 0, 1, 0, 0, 1]
UPS = [0.90, 1.62, 6, 0.62, 2, 3, 1, 1, 3]
prob.decisions = [jmoo_decision(names[i], LOWS[i], UPS[i]) for i in range(len(names))]
prob.objectives = [jmoo_objective("Cost", True), jmoo_objective("Score", False),
jmoo_objective("Completion", False), jmoo_objective("Idle", True)]
def __init__(prob, n):
prob.name = "Kursawe"
names = ["x1", "x2", "x3", "x4", "x5", "x6", "x7", "x8"]
n = max(min(8, n), 1) #constrain n to its range
prob.decisions = [jmoo_decision(names[i], -5, 5) for i in range(n)]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob):
prob.name = "ConstrEx"
prob.decisions = [
jmoo_decision("x1", 0.1, 1.0),
jmoo_decision("x2", 0, 5)
]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob):
prob.name = "ZDT6"
names = ["x" + str(i + 1) for i in range(10)]
prob.decisions = [
jmoo_decision(names[i], 0, 1) for i in range(len(names))
]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob):
prob.name = "Poloni"
prob.decisions = [
jmoo_decision("x1", -math.pi, math.pi),
jmoo_decision("x2", -math.pi, math.pi)
]
prob.objectives = [
jmoo_objective("f1", False),
jmoo_objective("f2", False)
]
def __init__(prob, n=30):
prob.name = "UF10"
prob.decisions = [
jmoo_decision("x" + str(i), 0, 1) for i in range(2)
] + [jmoo_decision("x" + str(j), -2, 2) for j in range(2, n)]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True),
jmoo_objective("f3", True)
]
def __init__(prob):
prob.name = "Tanaka"
prob.decisions = [
jmoo_decision("x1", -math.pi, math.pi),
jmoo_decision("x2", -math.pi, math.pi)
]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob):
prob.name = "POM3B"
names = ["Culture", "Criticality", "Criticality Modifier", "Initial Known", "Inter-Dependency",
"Dynamism", "Size", "Plan", "Team Size"]
LOWS = [0.10, 0.82, 80, 0.40, 0, 1, 0, 0, 1]
UPS = [0.90, 1.26, 95, 0.70, 100, 50, 2, 5, 20]
prob.decisions = [jmoo_decision(names[i], LOWS[i], UPS[i]) for i in range(len(names))]
prob.objectives = [jmoo_objective("Cost", True, 0), jmoo_objective("Score", True, 0, 1),
# jmoo_objective("Completion", False, 0, 1)]
jmoo_objective("Idle", True, 0, 1)]
def __init__(prob):
prob.name = "TwoBarTruss"
prob.decisions = [
jmoo_decision("x1", 0, 0.01),
jmoo_decision("x2", 0, 0.01),
jmoo_decision("x3", 1, 3)
]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob):
prob.name = "Viennet4"
prob.decisions = [
jmoo_decision("x1", -4, 4),
jmoo_decision("x2", -4, 4)
]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True),
jmoo_objective("f3", True)
]
def __init__(prob, percentage=-1):
prob.percentage=percentage
prob.name = "POM3D_1_2"
names = ["Culture", "Criticality", "Criticality Modifier", "Initial Known", "Inter-Dependency", "Dynamism",
"Size", "Plan", "Team Size"]
LOWS = [0.10, 0.82, 2, 0.60, 80, 1, 0, 0, 10]
UPS = [0.20, 1.26, 8, 0.95, 100, 10, 2, 5, 20]
prob.decisions = [jmoo_decision(names[i], LOWS[i], UPS[i]) for i in range(len(names))]
# prob.objectives = [jmoo_objective("Cost", True, 0), jmoo_objective("Score", False, 0, 1),
# jmoo_objective("Completion", False, 0, 1)]#, jmoo_objective("Idle", True, 0, 1)]
prob.objectives = [jmoo_objective("Cost", True, 0), jmoo_objective("Score", True, 0, 1)]
def __init__(prob, n):
prob.evals = 0
prob.name = "Fonseca"
names = ["x1", "x2", "x3", "x4", "x5", "x6", "x7", "x8"]
n = max(min(8, n), 1) #constrain n to its range
prob.decisions = [
jmoo_decision("x" + str(i + 1), -2, 2) for i in range(n)
]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob):
prob.name = "Osyczka2"
lows = [0, 0, 1, 1, 0, 0]
ups = [10, 10, 5, 5, 6, 10]
names = ["x1", "x2", "x3", "x4", "x5", "x6"]
prob.decisions = [
jmoo_decision(names[i], lows[i], ups[i]) for i in range(6)
]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(prob):
prob.name = "Golinski"
names = ["x1", "x2", "x3", "x4", "x5", "x6", "x7"]
LOWS = [2.6, 0.7, 17.0, 7.3, 7.3, 2.9, 5.0]
UPS = [3.6, 0.8, 28.0, 8.3, 8.3, 3.9, 5.5]
prob.decisions = [
jmoo_decision(names[i], LOWS[i], UPS[i]) for i in range(7)
]
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(self, name, objnum = 3):
self.name = name
assert(if_exists(name) is True), "Check the filename"
self.url = "./Problems/Feature_Models/References/" + name + ".xml"
spl_cost_data = "./Problems/Feature_Models/Cost/" + name + ".cost"
self.ft = load_ft_url(self.url)
self.ft.load_cost(spl_cost_data)
lows = [0 for _ in xrange(len(self.ft.leaves))]
ups = [1 for _ in xrange(len(self.ft.leaves))]
names = ["x"+str(i) for i in xrange(len(self.ft.leaves))]
self.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in xrange(len(self.ft.leaves))]
self.objectives = [jmoo_objective("fea", True), jmoo_objective("conVio", True), jmoo_objective("Cost", True)]
def __init__(prob):
prob.name = "ZDT4"
names = ["x" + str(i + 1) for i in range(10)]
prob.decisions = [
jmoo_decision(names[i], -5, 5) for i in range(len(names))
]
prob.decisions[0].low = 0
prob.decisions[0].up = 1
prob.objectives = [
jmoo_objective("f1", True),
jmoo_objective("f2", True)
]
def __init__(self, name, objnum = 3):
self.name = name
assert(if_exists(name) is True), "Check the filename"
self.url = "./Problems/Feature_Models/References/" + name + ".xml"
spl_cost_data = "./Problems/Feature_Models/Cost/" + name + ".cost"
self.ft = load_ft_url(self.url)
self.ft.load_cost(spl_cost_data)
lows = [0 for _ in xrange(len(self.ft.leaves))]
ups = [1 for _ in xrange(len(self.ft.leaves))]
names = ["x"+str(i) for i in xrange(len(self.ft.leaves))]
self.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in xrange(len(self.ft.leaves))]
self.objectives = [jmoo_objective("fea", True), jmoo_objective("conVio", True), jmoo_objective("Cost", True)]
def __init__(self, name, valid_solutions=False, objnum=3, is_binary=False):
self.name = name
self.valid_solutions = valid_solutions
assert(if_exists(name) is True), "Check the filename"
self.url = "./Problems/Feature_Models/References/" + name + ".xml"
spl_cost_data = "./Problems/Feature_Models/Cost/" + name + ".cost"
self.ft = load_ft_url(self.url)
self.ft.load_cost(spl_cost_data)
lows = [0 for _ in xrange(len(self.ft.leaves))]
ups = [1 for _ in xrange(len(self.ft.leaves))]
names = ["x"+str(i) for i in xrange(len(self.ft.leaves))]
self.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in xrange(len(self.ft.leaves))]
self.objectives = [jmoo_objective("number_of_features", True),
jmoo_objective("constrained_violated", True),
jmoo_objective("cost", True)]
self.is_binary = True
def __init__(self, treatment, requirements=9, name="CPM_APACHE", filename="./data/Apache_AllMeasurements.csv"):
# def __init__(self, treatment, number=50, requirements=9, name="CPM_APACHE", filename="./Problems/CPM/data/Apache_AllMeasurements.csv"):
self.name = name
self.filename = filename
self.no_of_clusters = 0
# Setting up to create decisions
names = ["x"+str(i+1) for i in xrange(requirements)]
lows = [0 for _ in xrange(requirements)]
ups = [1 for _ in xrange(requirements)]
# Generating decisions
self.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(requirements)]
# Generating Objectives (this is single objective)
self.objectives = [jmoo_objective("f1", True)]
# Read data
self.header, self.data = read_csv(self.filename, header=True)
self.training_independent, self.training_dependent, = self.get_training_data(method=treatment)
global training_percent
from math import log, ceil
# # print training_percent,
# print "=" * 20
# print "Reduced data: ", self.training_dependent
# print "total run time: ", sum(self.training_dependent)
# print "totol total run time: ", self.find_total_time()
# print "sadsadsa time: ", self.find_total_time() - sum(self.training_dependent)
# print "Saving Percentage: ", (sum(self.training_dependent)/self.find_total_time()) *100
# print "Length of self.data: ", len(self.data)
# print treatment.__name__
print "Length of training dataset: ", len(self.training_dependent), len(self.data), (2*log(len(self.data) * training_percent, 2))
self.CART = tree.DecisionTreeRegressor()
self.CART = self.CART.fit(self.training_independent, self.training_dependent)
self.saved_time = (self.find_total_time() - sum(self.training_dependent))/10**4
def __init__(self, treatment, requirements=9, name="CPM_APACHE", filename="./Data/Apache_AllMeasurements.csv"):
# def __init__(self, treatment, number=50, requirements=9, name="CPM_APACHE", filename="./Problems/CPM/Data/Apache_AllMeasurements.csv"):
self.name = name
self.filename = filename
self.no_of_clusters = 0
# Setting up to create decisions
names = ["x"+str(i+1) for i in xrange(requirements)]
lows = [0 for _ in xrange(requirements)]
ups = [1 for _ in xrange(requirements)]
# Generating decisions
self.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(requirements)]
# Generating Objectives (this is single objective)
self.objectives = [jmoo_objective("f1", True)]
# Read Data
self.header, self.data = read_csv(self.filename, header=True)
self.training_independent, self.training_dependent, = self.get_training_data(method=treatment)
global training_percent
from math import log, ceil
# # print training_percent,
# print "=" * 20
# print "Reduced Data: ", self.training_dependent
# print "total run time: ", sum(self.training_dependent)
# print "totol total run time: ", self.find_total_time()
# print "sadsadsa time: ", self.find_total_time() - sum(self.training_dependent)
# print "Saving Percentage: ", (sum(self.training_dependent)/self.find_total_time()) *100
# print "Length of self.Data: ", len(self.Data)
# print treatment.__name__
print "Length of training dataset: ", len(self.training_dependent), len(self.data), (2*log(len(self.data) * training_percent, 2))
self.CART = tree.DecisionTreeRegressor()
self.CART = self.CART.fit(self.training_independent, self.training_dependent)
self.saved_time = (self.find_total_time() - sum(self.training_dependent))/10**4
def __init__(self, treatment, number=50, requirements=16, fraction=0.5, name="cpm_X264", filename="./Data/X264_AllMeasurements.csv"):
# def __init__(self, treatment, number=50, requirements=16, fraction=0.5, name="cpm_X264", filename="./Problems/CPM/Data/X264_AllMeasurements.csv"):
self.name = name
self.filename = filename
# Setting up to create decisions
names = ["x"+str(i+1) for i in xrange(requirements)]
lows = [0 for _ in xrange(requirements)]
ups = [1 for _ in xrange(requirements)]
# Generating decisions
self.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(requirements)]
# Generating Objectives (this is single objective)
self.objectives = [jmoo_objective("f1", True)]
# Read Data
self.header, self.data = read_csv(self.filename, header=True)
self.training_independent, self.training_dependent = self.get_training_data(method=treatment)
global training_percent
# print training_percent,
from math import log
# print "Length of training dataset: ", len(self.training_dependent), len(self.Data), (2*log(len(self.Data) * training_percent, 2))
self.CART = tree.DecisionTreeRegressor()
self.CART = self.CART.fit(self.training_independent, self.training_dependent)
self.saved_time = (self.find_total_time() - sum(self.training_dependent))/10**4
def __init__(self,
treatment,
number=50,
requirements=16,
fraction=0.5,
name="cpm_X264",
filename="./Data/X264_AllMeasurements.csv"):
self.name = name
self.filename = filename
if treatment is None:
treatment = random_where
elif treatment == 0:
treatment = base_line
# Setting up to create decisions
names = ["x" + str(i + 1) for i in xrange(requirements)]
lows = [0 for _ in xrange(requirements)]
ups = [1 for _ in xrange(requirements)]
# Generating decisions
self.decisions = [
jmoo_decision(names[i], lows[i], ups[i])
for i in range(requirements)
]
# Generating Objectives (this is single objective)
self.objectives = [jmoo_objective("f1", True)]
# Read Data
self.header, self.data = read_csv(self.filename, header=True)
self.training_independent, self.training_dependent = self.get_training_data(
method=treatment)
self.CART = tree.DecisionTreeRegressor()
self.CART = self.CART.fit(self.training_independent,
self.training_dependent)
self.saved_time = (self.find_total_time() -
sum(self.training_dependent)) / 10**4
def __init__(prob, numDecs=20, numObjs=2):
prob.name = "DTLZ6_" + str(numDecs) + "_" + str(numObjs)
names = ["x"+str(i+1) for i in range(numDecs)]
lows = [0.0 for i in range(numDecs)]
ups = [1.0 for i in range(numDecs)]
prob.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(numDecs)]
prob.objectives = [jmoo_objective("f" + str(i+1), True) for i in range(numObjs)]
def __init__(
self,
treatment,
number=50,
requirements=16,
fraction=0.5,
name="cpm_X264",
filename="./Data/X264_AllMeasurements.csv",
):
# def __init__(self, treatment, number=50, requirements=16, fraction=0.5, name="cpm_X264", filename="./Problems/CPM/Data/X264_AllMeasurements.csv"):
self.name = name
self.filename = filename
# Setting up to create decisions
names = ["x" + str(i + 1) for i in xrange(requirements)]
lows = [0 for _ in xrange(requirements)]
ups = [1 for _ in xrange(requirements)]
# Generating decisions
self.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(requirements)]
# Generating Objectives (this is single objective)
self.objectives = [jmoo_objective("f1", True)]
# Read Data
self.header, self.data = read_csv(self.filename, header=True)
self.training_independent, self.training_dependent = self.get_training_data(method=treatment)
global training_percent
# print training_percent,
from math import log
# print "Length of training dataset: ", len(self.training_dependent), len(self.Data), (2*log(len(self.Data) * training_percent, 2))
self.CART = tree.DecisionTreeRegressor()
self.CART = self.CART.fit(self.training_independent, self.training_dependent)
self.saved_time = (self.find_total_time() - sum(self.training_dependent)) / 10 ** 4
def __init__(self, requirements, releases, clients, density, budget):
self.name = (
"NRP_"
+ str(requirements)
+ "_"
+ str(releases)
+ "_"
+ str(clients)
+ "_"
+ str(density)
+ "_"
+ str(budget)
)
names = ["x" + str(i + 1) for i in range(requirements)] # |x_i + y_i|
lows = [-1 for i in xrange(requirements)]
ups = [(releases - 1) for _ in xrange(requirements)]
self.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(requirements)]
self.objectives = [jmoo_objective("f1", False)] # , jmoo_objective("f2", False)] # single objective nrp
self.trequirements = requirements
self.treleases = releases
self.tclients = clients
self.tdensity = density
self.tbudget = budget
self.requirement = None
self.client = None
self.release = None
self.precedence = []
self.generate_data()
def __init__(
self, treatment, number=50, requirements=18, name="CPM_BDBC", filename="./Data/BDBC_AllMeasurements.csv"
):
# def __init__(self, treatment, number=50, requirements=18, name="CPM_BDBC", filename="./Problems/CPM//Data/BDBC_AllMeasurements.csv"):
self.name = name
self.filename = filename
if treatment is None:
treatment = east_west_where
elif treatment == 0:
treatment = base_line
names = ["x" + str(i + 1) for i in xrange(requirements)]
lows = [0 for _ in xrange(requirements)]
ups = [1 for _ in xrange(requirements)]
self.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(requirements)]
self.objectives = [jmoo_objective("f1", True)]
self.header, self.data = read_csv(self.filename, header=True)
self.training_independent, self.training_dependent = self.get_training_data(method=treatment)
global training_percent
# print "inside: ", training_percent
from math import log
# print "Length of training dataset: ", len(self.training_dependent), len(self.Data), (2*log(len(self.Data) * training_percent, 2))
self.CART = tree.DecisionTreeRegressor()
self.CART = self.CART.fit(self.training_independent, self.training_dependent)
self.saved_time = (self.find_total_time() - sum(self.training_dependent)) / 10 ** 4
def __init__(self, treatment, requirements=9, name="CPM_APACHE", filename="./Data/Apache_AllMeasurements.csv"):
self.name = name
self.filename = filename
if treatment is None:
treatment = random_where
elif treatment == 0:
treatment = base_line
# Setting up to create decisions (This is something specific from the JMOO framework
names = ["x" + str(i + 1) for i in xrange(requirements)]
lows = [0 for _ in xrange(requirements)]
ups = [1 for _ in xrange(requirements)]
# Generating decisions
self.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(requirements)]
# Generating Objectives (this is single objective)
self.objectives = [jmoo_objective("f1", True)]
# Read Data
self.header, self.data = read_csv(self.filename, header=True)
self.training_independent, self.training_dependent = self.get_training_data(method=treatment)
self.CART = tree.DecisionTreeRegressor()
self.CART = self.CART.fit(self.training_independent, self.training_dependent)
self.saved_time = (self.find_total_time() - sum(self.training_dependent))/10**4
def __init__(prob):
prob.name = "POM3Csanscomp"
names = [
"Culture", "Criticality", "Criticality Modifier", "Initial Known",
"Inter-Dependency", "Dynamism", "Size", "Plan", "Team Size"
]
LOWS = [0.50, 0.82, 2, 0.20, 0, 40, 2, 0, 20]
UPS = [0.90, 1.26, 8, 0.50, 50, 50, 4, 5, 44]
prob.decisions = [
jmoo_decision(names[i], LOWS[i], UPS[i]) for i in range(len(names))
]
prob.objectives = [
jmoo_objective("Cost", True, 0),
jmoo_objective("Score", False, 0, 1),
jmoo_objective("Idle", True, 0, 1)
]
def __init__(self,
treatment,
number=50,
requirements=18,
name="CPM_BDBC",
filename="./Data/BDBC_AllMeasurements.csv"):
# def __init__(self, treatment, number=50, requirements=18, name="CPM_BDBC", filename="./Problems/CPM//Data/BDBC_AllMeasurements.csv"):
self.name = name
self.filename = filename
if treatment is None: treatment = east_west_where
elif treatment == 0: treatment = base_line
names = ["x" + str(i + 1) for i in xrange(requirements)]
lows = [0 for _ in xrange(requirements)]
ups = [1 for _ in xrange(requirements)]
self.decisions = [
jmoo_decision(names[i], lows[i], ups[i])
for i in range(requirements)
]
self.objectives = [jmoo_objective("f1", True)]
self.header, self.data = read_csv(self.filename, header=True)
self.training_independent, self.training_dependent = self.get_training_data(
method=treatment)
global training_percent
# print "inside: ", training_percent
from math import log
# print "Length of training dataset: ", len(self.training_dependent), len(self.Data), (2*log(len(self.Data) * training_percent, 2))
self.CART = tree.DecisionTreeRegressor()
self.CART = self.CART.fit(self.training_independent,
self.training_dependent)
self.saved_time = (self.find_total_time() -
sum(self.training_dependent)) / 10**4
def __init__(prob, percentage=-1):
prob.percentage = percentage
prob.name = "POM3B_2_3"
names = [
"Culture", "Criticality", "Criticality Modifier", "Initial Known",
"Inter-Dependency", "Dynamism", "Size", "Plan", "Team Size"
]
LOWS = [0.10, 0.82, 80, 0.40, 0, 1, 0, 0, 1]
UPS = [0.90, 1.26, 95, 0.70, 100, 50, 2, 5, 20]
prob.decisions = [
jmoo_decision(names[i], LOWS[i], UPS[i]) for i in range(len(names))
]
prob.objectives = [
jmoo_objective("Score", True, 0, 1),
jmoo_objective("Idle", True, None, None)
]
def __init__(prob, numDecs=10, numObjs=3):
super(car_impact, prob).__init__()
prob.name = "Car_Impact_" + str(numDecs) + "_" + str(numObjs)
names = ["x"+str(i+1) for i in range(numDecs)]
lows = [0.0 for i in range(numDecs)]
ups = [1.0 for i in range(numDecs)]
prob.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(numDecs)]
prob.objectives = [jmoo_objective("f" + str(i+1), True) for i in range(numObjs)]
def __init__(prob, numDecs=10, numObjs=2):
super(dtlz5, prob).__init__()
prob.name = "DTLZ5"
names = ["x"+str(i+1) for i in range(numDecs)]
lows = [0.0 for i in range(numDecs)]
ups = [1.0 for i in range(numDecs)]
prob.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(numDecs)]
prob.objectives = [jmoo_objective("f" + str(i+1), True) for i in range(numObjs)]
def __init__(prob, numDecs=10, numObjs=2):
super(c2_convex_dtlz2, prob).__init__()
prob.name = "Convex_DTLZ2_" + str(numDecs) + "_" + str(numObjs)
names = ["x"+str(i+1) for i in range(numDecs)]
lows = [0.0 for i in range(numDecs)]
ups = [1.0 for i in range(numDecs)]
prob.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(numDecs)]
prob.objectives = [jmoo_objective("f" + str(i+1), True) for i in range(numObjs)]
def __init__(prob, dataset_name, instances=100, features=20, nol=30, noc=0.3, tuning_precent=20):
prob.name = "hpcc_kmeans_" + dataset_name
prob.features = features
prob.instances = instances
prob.dataset_name = dataset_name
prob.tuning_instances = sample(range(1, prob.instances + 1), int(prob.instances * tuning_precent / 100))
prob.conv = noc
prob.decisions = [jmoo_decision("k", 1, int(features ** 0.5)), jmoo_decision("number_of_loops", 1, nol)]
prob.objectives = [jmoo_objective("convergence", True)]
prob.is_binary = False
def __init__(prob, dataset_name, features=20):
prob.name = "hpcc_rf_" + dataset_name
prob.dataset_name = dataset_name
prob.decisions = [
jmoo_decision("no_trees", 80, 120),
jmoo_decision("no_features", 4, features),
jmoo_decision("purity", 0.6, 1),
jmoo_decision("depth", 80, 120),
]
prob.objectives = [jmoo_objective("accuracy", True)]
prob.is_binary = False
def __init__(self, requirements=18, fraction=0.5, name="CPM_BDBC", filename="./Problems/CPM/Data/BDBC_AllMeasurements.csv"):
self.name = name
self.filename = filename
names = ["x"+str(i+1) for i in xrange(requirements)]
lows = [0 for _ in xrange(requirements)]
ups = [1 for _ in xrange(requirements)]
self.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(requirements)]
self.objectives = [jmoo_objective("f1", True)]
self.data = read_csv(self.filename)
self.testing_independent, self.testing_dependent = [], []
self.training_independent, self.training_dependent = self.get_training_data(fraction)
self.CART = tree.DecisionTreeRegressor()
self.CART = self.CART.fit(self.training_independent, self.training_dependent)
def __init__(self, tree, obj_funcs=None, **settings):
jmoo_problem.__init__(self)
if not obj_funcs:
obj_funcs = [eval_softgoals, eval_goals, eval_coverage]
self.name = tree.name
self.obj_funcs = obj_funcs
self._tree = tree
self.roots = self._tree.get_roots()
self.bases = self._tree.get_bases()
obj_names = [func.__name__.split("_")[1] for func in obj_funcs]
dec_names = [base.id for base in self.bases]
self.decisions = [jmoo_decision(dec_names[i], f, t) for i in range(len(dec_names))]
self.objectives = [jmoo_objective(obj_names[i], True) for i in range(len(obj_names))]
self.chain = set()
self.is_percent = settings.get("obj_is_percent", True)
def __init__(self, treatment, number=50, requirements=18, name="CPM_BDBC", filename="./Problems/CPM//Data/BDBC_AllMeasurements.csv"):
self.name = name
self.filename = filename
if treatment is None: treatment = random_where
elif treatment == 0: treatment = base_line
names = ["x"+str(i+1) for i in xrange(requirements)]
lows = [0 for _ in xrange(requirements)]
ups = [1 for _ in xrange(requirements)]
self.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(requirements)]
self.objectives = [jmoo_objective("f1", True)]
self.header, self.data = read_csv(self.filename, header=True)
print "Length of Data: ", len(self.data)
self.training_independent, self.training_dependent = self.get_training_data(method=treatment)
self.CART = tree.DecisionTreeRegressor()
self.CART = self.CART.fit(self.training_independent, self.training_dependent)
def __init__(self, treatment, number=50, requirements=16, fraction=0.5, name="cpm_X264", filename="./Problems/CPM/Data/X264_AllMeasurements.csv"):
self.name = name
self.filename = filename
if treatment is None: treatment = random_where
elif treatment == 0: treatment = base_line
# Setting up to create decisions
names = ["x"+str(i+1) for i in xrange(requirements)]
lows = [0 for _ in xrange(requirements)]
ups = [1 for _ in xrange(requirements)]
# Generating decisions
self.decisions = [jmoo_decision(names[i], lows[i], ups[i]) for i in range(requirements)]
# Generating Objectives (this is single objective)
self.objectives = [jmoo_objective("f1", True)]
# Read Data
self.header, self.data = read_csv(self.filename, header=True)
self.training_independent, self.training_dependent = self.get_training_data(method=treatment)
self.CART = tree.DecisionTreeRegressor()
self.CART = self.CART.fit(self.training_independent, self.training_dependent)
def __init__(prob, n=30):
prob.name = "UF10"
prob.decisions = [jmoo_decision("x" + str(i), 0, 1) for i in range(2)] + [jmoo_decision("x" + str(j), -2, 2) for j in range(2, n)]
prob.objectives = [jmoo_objective("f1", True), jmoo_objective("f2", True), jmoo_objective("f3", True)]
def __init__(prob, n):
prob.name = "JoeTest1"
names = ["x1", "x2", "x3"]
prob.decisions = [jmoo_decision(names[i], -2, 2) for i in range(n)]
prob.objectives = [jmoo_objective("f1", True), jmoo_objective("f2", True)]
def __init__(prob):
prob.name = "Tanaka"
prob.decisions = [jmoo_decision("x1",-math.pi, math.pi),jmoo_decision("x2", -math.pi, math.pi)]
prob.objectives = [jmoo_objective("f1", True), jmoo_objective("f2", True)]
def __init__(prob):
prob.name = "ConstrEx"
prob.decisions = [jmoo_decision("x1",0.1, 1.0),jmoo_decision("x2", 0, 5)]
prob.objectives = [jmoo_objective("f1", True), jmoo_objective("f2", True)]
def __init__(prob):
prob.name = "Srinivas"
prob.decisions = [jmoo_decision("x" + str(i+1), -20, 20) for i in range(2)]
prob.objectives = [jmoo_objective("f1", True), jmoo_objective("f2", True)]