def setUp(self):
# model a -> b
bn = {}
self.arr = []
bn[0] = BN(domain=Domain(),
new_domain_variables={
'a': [0, 1],
'b': [0, 1]
})
bn[0].add_cpts([
CPT(Factor(variables=['a'], data=[0.5, 0.5]), child='a'),
CPT(Factor(variables=['a', 'b'], data=[0.3, 0.7, 0.4, 0.6]),
child='b')
])
self.arr.append([('a', 'b')])
bn[1] = BN(domain=Domain(),
new_domain_variables={
'a': [0, 1],
'b': [0, 1],
'c': [0, 1]
})
bn[1].add_cpts([
CPT(Factor(variables=['a'], data=[0.5, 0.5]), child='a'),
CPT(Factor(variables=['a', 'b'], data=[0.3, 0.7, 0.4, 0.6]),
child='b'),
CPT(Factor(variables=['c', 'b'], data=[0.1, 0.9, 0.2, 0.8]),
child='c')
])
self.arr.append([('a', 'b'), ('b', 'c')])
self.cbn = [CBN.from_bn(bn[i]) for i in bn.keys()]
def testdnet(self):
from gPy.IO import read_dnet
from gPy.Models import BN
from gPy.Variables import Domain
bnm = BN(domain=Domain())
bnm.from_dnet(read_dnet('Asia.dnet'))
self.samegraph(bnm.adg(),self.asia_adg)
for name, cpt_in_file in self.asia_cpts.items():
cpt = bnm[name]
self.samecpt(cpt,cpt_in_file,cpt.child())
def testdnet(self):
from gPy.IO import read_dnet
from gPy.Models import BN
from gPy.Variables import Domain
bnm = BN(domain=Domain())
bnm.from_dnet(read_dnet('Asia.dnet'))
self.samegraph(bnm.adg(), self.asia_adg)
for name, cpt_in_file in self.asia_cpts.items():
cpt = bnm[name]
self.samecpt(cpt, cpt_in_file, cpt.child())
def rand_bn(vs, max_potential_parents=15):
model = BN(domain=Domain(), new_domain_variables=vs)
for child in vs.keys():
parents = list(model.variables())
too_many = len(parents) - max_potential_parents
if too_many > 0:
for i in xrange(too_many):
parents.remove(choice(parents))
fv = rand_subset(parents) | set([child])
n = reduce(operator.mul, [len(vs[v]) for v in fv])
f = Factor(variables=fv,
data=rand_factor_data(n),
domain=model,
check=True)
cpt = CPT(f, child, True, True)
model *= cpt
return model
def rand_bn(vs, max_potential_parents = 15):
model = BN(domain = Domain(), new_domain_variables = vs)
for child in vs.keys():
parents = list(model.variables())
too_many = len(parents) - max_potential_parents
if too_many > 0:
for i in xrange(too_many):
parents.remove(choice(parents))
fv = rand_subset(parents) | set([child])
n = reduce(operator.mul, [len(vs[v]) for v in fv])
f = Factor(variables = fv
,data = rand_factor_data(n)
,domain = model
,check = True
)
cpt = CPT(f,child,True,True)
model *= cpt
return model
def setUp(self):
from gPy.Variables import Domain
bnm = BN(domain=Domain())
bnm.from_dnet(read_dnet('Asia.dnet'))
self.hypergraph = bnm._hypergraph
self.adg = bnm._adg
self.tarjan = UGraph(range(1,11),
((1,2),(1,3),(2,3),(2,10),(3,10),(4,5),
(4,7),(5,6),(5,9),(5,7),(6,7),(6,9),
(7,8),(7,9),(8,9),(8,10),(9,10)))
self.tarjan2 = UGraph(range(1,10),
((1,4),(1,3),(2,3),(2,7),(3,5),(3,6),
(4,5),(4,8),(5,6),(5,8),(6,7),(6,9),
(7,9),(8,9)))
self.tarjan3 = UGraph(range(1,10),
((1,4),(1,3),(2,3),(2,7),(3,5),(3,6),
(4,5),(4,8),(5,6),(5,8),(6,7),(6,9),
(7,9),(8,9),
(3,4),(3,7),(4,6),(4,7),(5,7),(6,8),(7,8)))
self.tarjanh1 = Hypergraph([[3,4],[2,4],[1,2,3]])
self.tarjanh2 = Hypergraph([[3,4],[2,4],[1,2,3],[2,3,4]])
self.graph1 = UGraph('ABCDEF',('AB','AC','BD','CE','EF'))
self.graph2 = UGraph('ABCDEF',('AB','AC','BD','CE','EF','BC','CD','DE'))
def setUp(self):
from gPy.Variables import Domain
self.domain = Domain()
self.bnm = BN(domain=self.domain)
self.bnm.from_dnet(read_dnet('Asia.dnet'))
self.cptdict = {}
# taken directly from Netica output
self.marginals = [
Factor((('VisitAsia'),),
[0.99,0.01]),
Factor((('Tuberculosis'),),
[0.9896,0.0104]),
Factor((('Smoking'),),
[0.5,0.5]),
Factor((('Cancer'),),
[0.945,0.055]),
Factor((('TbOrCa'),),
[0.93517, 0.064828]),
Factor((('XRay'),),
[0.11029, 0.88971]),
Factor((('Bronchitis'),),
[0.55,0.45]),
Factor((('Dyspnea'),),
[0.56403,0.43597])
]
# taken directly from Netica output
self.cond_marginals = [
Factor((('VisitAsia'),),
[0.95192,0.048077]),
Factor((('Tuberculosis'),),
[0,1]),
Factor((('Smoking'),),
[0.52381,0.47619]),
#other marginals are conditional on these values
#Factor((('Cancer'),),
# [1,0]),
#Factor((('TbOrCa'),),
# [0,1]),
Factor((('XRay'),),
[0.98, 0.02]),
Factor((('Bronchitis'),),
[0.55714,0.44286]),
Factor((('Dyspnea'),),
[0.21143,0.78857])
]
for cpt in self.bnm:
self.cptdict[cpt.child()] = cpt
self.rawdata = read_csv(open('alarm_1K.dat'))
def setUp(self):
from gPy.Variables import Domain
self.domain = Domain()
self.bnm = BN(domain=self.domain)
self.bnm.from_dnet(read_dnet('Asia.dnet'))
self.cptdict = {}
# taken directly from Netica output
self.marginals = [
Factor((('VisitAsia'), ), [0.99, 0.01]),
Factor((('Tuberculosis'), ), [0.9896, 0.0104]),
Factor((('Smoking'), ), [0.5, 0.5]),
Factor((('Cancer'), ), [0.945, 0.055]),
Factor((('TbOrCa'), ), [0.93517, 0.064828]),
Factor((('XRay'), ), [0.11029, 0.88971]),
Factor((('Bronchitis'), ), [0.55, 0.45]),
Factor((('Dyspnea'), ), [0.56403, 0.43597])
]
# taken directly from Netica output
self.cond_marginals = [
Factor((('VisitAsia'), ), [0.95192, 0.048077]),
Factor((('Tuberculosis'), ), [0, 1]),
Factor((('Smoking'), ), [0.52381, 0.47619]),
#other marginals are conditional on these values
#Factor((('Cancer'),),
# [1,0]),
#Factor((('TbOrCa'),),
# [0,1]),
Factor((('XRay'), ), [0.98, 0.02]),
Factor((('Bronchitis'), ), [0.55714, 0.44286]),
Factor((('Dyspnea'), ), [0.21143, 0.78857])
]
for cpt in self.bnm:
self.cptdict[cpt.child()] = cpt
self.rawdata = read_csv(open('alarm_1K.dat'))
def generate_dense_bn(density, num_vars=8, num_vals=3):
if density > num_vars:
raise RuntimeError, 'density must be less than number of variables'
vars, parents = generate_dense_parents(density, num_vars)
vals = dict([(var, frozenset([i for i in xrange(num_vals)]))
for var in vars])
bn = BN(domain=Domain(), new_domain_variables=vals)
for child in vars:
if child in parents:
n = num_vals**(len(parents[child]) + 1)
else:
n = num_vals
parents[child] = frozenset()
f = Factor(variables=frozenset([child]) | parents[child],
data=rand_factor_data(n),
domain=bn,
check=True)
bn *= CPT(f, child, True, True)
return bn
from gPy.Examples import minibn, asia
from gPy.Models import FR,BN
from gPy.Parameters import Factor,CPT
from gPy.Variables import Domain
from random import choice,randrange,uniform,shuffle
import operator, unittest, pickle
xor = BN(domain=Domain(), new_domain_variables={'a': [0,1], 'b':[0,1], 'c':[0,1]})
xor.add_cpts([CPT(Factor(variables=['a'], data=[0.5, 0.5]),child='a')
,CPT(Factor(variables=['b'], data=[0.5, 0.5]),child='b')
,CPT(Factor(variables=['c','a','b'], data=[1.0,0.0,0.0,1.0,0.0,1.0,1.0,0.0]),child='c')
])
cbn_small_names = ['xor','minibn','asia']
cbn_small_test_cases = [xor,minibn,asia]
cbn_large_names = ['alarm','insurance','carpo']
try:
# load the pickled large Bayes nets.
cbn_large_test_cases = map(lambda fn: pickle.load(open('networks/'+fn+'_bn.pck','r')),
cbn_large_names)
except:
cbn_large_names = []
cbn_large_test_cases = []
cbn_test_cases = cbn_small_test_cases + cbn_large_test_cases
def distribution_of(model):
"""Returns a normalised factor representing the joint instantiation
of the model.
"""
from gPy.Examples import minibn, asia
from gPy.Models import FR, BN
from gPy.Parameters import Factor, CPT
from gPy.Variables import Domain
from random import choice, randrange, uniform, shuffle
import operator, unittest, pickle
xor = BN(domain=Domain(),
new_domain_variables={
'a': [0, 1],
'b': [0, 1],
'c': [0, 1]
})
xor.add_cpts([
CPT(Factor(variables=['a'], data=[0.5, 0.5]), child='a'),
CPT(Factor(variables=['b'], data=[0.5, 0.5]), child='b'),
CPT(Factor(variables=['c', 'a', 'b'],
data=[1.0, 0.0, 0.0, 1.0, 0.0, 1.0, 1.0, 0.0]),
child='c')
])
cbn_small_names = ['xor', 'minibn', 'asia']
cbn_small_test_cases = [xor, minibn, asia]
cbn_large_names = ['alarm', 'insurance', 'carpo']
try:
# load the pickled large Bayes nets.
cbn_large_test_cases = map(
lambda fn: pickle.load(open('networks/' + fn + '_bn.pck', 'r')),
cbn_large_names)
except:
cbn_large_names = []
cbn_large_test_cases = []
class TestParameters(unittest.TestCase):
def setUp(self):
from gPy.Variables import Domain
self.domain = Domain()
self.bnm = BN(domain=self.domain)
self.bnm.from_dnet(read_dnet('Asia.dnet'))
self.cptdict = {}
# taken directly from Netica output
self.marginals = [
Factor((('VisitAsia'), ), [0.99, 0.01]),
Factor((('Tuberculosis'), ), [0.9896, 0.0104]),
Factor((('Smoking'), ), [0.5, 0.5]),
Factor((('Cancer'), ), [0.945, 0.055]),
Factor((('TbOrCa'), ), [0.93517, 0.064828]),
Factor((('XRay'), ), [0.11029, 0.88971]),
Factor((('Bronchitis'), ), [0.55, 0.45]),
Factor((('Dyspnea'), ), [0.56403, 0.43597])
]
# taken directly from Netica output
self.cond_marginals = [
Factor((('VisitAsia'), ), [0.95192, 0.048077]),
Factor((('Tuberculosis'), ), [0, 1]),
Factor((('Smoking'), ), [0.52381, 0.47619]),
#other marginals are conditional on these values
#Factor((('Cancer'),),
# [1,0]),
#Factor((('TbOrCa'),),
# [0,1]),
Factor((('XRay'), ), [0.98, 0.02]),
Factor((('Bronchitis'), ), [0.55714, 0.44286]),
Factor((('Dyspnea'), ), [0.21143, 0.78857])
]
for cpt in self.bnm:
self.cptdict[cpt.child()] = cpt
self.rawdata = read_csv(open('alarm_1K.dat'))
def test_str(self):
out = ''
for node in [
'VisitAsia', 'Tuberculosis', 'Smoking', 'Cancer', 'TbOrCa',
'XRay', 'Bronchitis', 'Dyspnea'
]:
out += str(self.cptdict[node])
self.assertEqual(out, open('Asia.cpts').read())
def test_cptcheck(self):
# be nice to check that this is printed out
errmsg = """
For child: Dyspnea
For row: Bronchitis=Absent, Dyspnea=Absent, TbOrCa=True
Sum was: 0.90 (should be 1.0)
"""
def bnfromfile(filename):
x = BN(domain=self.domain)
x.from_dnet(read_dnet(filename))
return x
self.assertRaises(CPTError, bnfromfile, open('Asia_wrong.dnet'))
def test_multinplace(self):
factor = self.cptdict['VisitAsia'] * self.cptdict['Tuberculosis']
fid = id(factor)
factor *= self.cptdict['Bronchitis']
self.assertEqual(fid, id(factor))
factor2 = self.cptdict['VisitAsia'] * \
self.cptdict['Tuberculosis'] * self.cptdict['Bronchitis']
self.samefactor(factor, factor2)
def test_scalarprod(self):
for factor in self.bnm:
for scalar in 0, 0.3, 7:
tf1 = scalar * factor
tf2 = factor * scalar
self.samefactor(tf1, tf2)
for i, val in enumerate(tf1._data):
self.assertAlmostEqual(val, scalar * factor._data[i],
places)
def test_instbyname(self):
dyspnea = self.cptdict['Dyspnea']
datum = dyspnea[{
'Dyspnea': 'Present',
'Bronchitis': 'Present',
'TbOrCa': 'False'
}]
self.assertAlmostEqual(datum, 0.8, places)
def test_getitem(self):
dyspnea = self.cptdict['Dyspnea']
datum = dyspnea['Present', 'Present', 'False']
self.assertAlmostEqual(datum, 0.8, places)
def test_z(self):
factor = 1
order = [
'VisitAsia', 'Tuberculosis', 'Smoking', 'Cancer', 'TbOrCa', 'XRay',
'Bronchitis', 'Dyspnea'
]
for name in order:
newfactor = factor * self.cptdict[name]
factor *= self.cptdict[name]
self.samefactor(factor, newfactor)
self.assertAlmostEqual(factor.z(), 1, places)
self.assertEqual(str(factor), open('Asia.joint').read())
factor = 1
for cpt in self.cptdict.values():
factor *= cpt
self.assertAlmostEqual(factor.z(), 1, places)
def test_marginals(self):
joint = 1
for cpt in self.bnm:
joint *= cpt
order = [
'VisitAsia', 'Tuberculosis', 'Smoking', 'Cancer', 'TbOrCa', 'XRay',
'Bronchitis', 'Dyspnea'
]
varset = set(order)
for i, var in enumerate(order):
tmp = varset.copy()
tmp.remove(var)
self.samefactor(self.marginals[i], joint.sumout(tmp))
def test_factorprodcommute(self):
for f1 in self.bnm:
for f2 in self.bnm:
tf1 = f1 * f2
tf2 = f2 * f1
self.samefactor(tf1, tf2)
def test_restrict(self):
bnm = self.bnm.copy(copy_domain=True)
given = {'Cancer': ['Absent'], 'TbOrCa': ['True']}
bnm.condition(given)
joint = 1
for cpt in bnm:
joint *= cpt
order = [
'VisitAsia', 'Tuberculosis', 'Smoking', 'XRay', 'Bronchitis',
'Dyspnea'
]
varset = set(order)
varset.update(['Cancer', 'TbOrCa'])
for i, var in enumerate(order):
marginal = joint.sumout(varset - set([var]))
marginal /= marginal.z()
self.samefactor(self.cond_marginals[i], marginal)
def samefactor(self, tf1, tf2):
self.assertEqual(tf1.variables(), tf2.variables())
tf2data = tf2.data()
for i, val in enumerate(tf1.data()):
self.assertAlmostEqual(val, tf2data[i], places)
def disp(fn, samples):
f = open(fn, 'w')
fact = samples.makeFactor(samples.variables())
for var in fact.variables():
print >> f, var,
print >> f, 'count'
for inst in fact.insts():
for i in inst:
print >> f, i,
print >> f, fact[inst]
f.close()
bn0 = BN(domain=Domain(), new_domain_variables={'a': [0, 1], 'b': [0, 1]})
bn0.add_cpts([
CPT(Factor(variables=['a'], data=[0.5, 0.5]), child='a'),
CPT(Factor(variables=['a', 'b'], data=[0.3, 0.7, 0.4, 0.6]), child='b')
])
w = CausalWorld(bn0)
samples = w.observe(10000)
disp('two_depend', samples)
bn1 = BN(domain=Domain(), new_domain_variables={'a': [0, 1], 'b': [0, 1]})
bn1.add_cpts([
CPT(Factor(variables=['a'], data=[0.5, 0.5]), child='a'),
CPT(Factor(variables=['b'], data=[0.3, 0.7]), child='b')
])
w = CausalWorld(bn1)
samples = w.observe(10000)
def bnfromfile(filename):
x = BN(domain=self.domain)
x.from_dnet(read_dnet(filename))
return x
from gPy.Variables import Domain
from gPy.LearningUtils import CausalWorld
def disp(fn, samples):
f = open(fn, 'w')
fact = samples.makeFactor(samples.variables())
for var in fact.variables():
print >>f, var,
print >>f, 'count'
for inst in fact.insts():
for i in inst:
print >>f, i,
print >>f, fact[inst]
f.close()
bn0 = BN(domain=Domain(), new_domain_variables={'a': [0,1], 'b':[0,1]})
bn0.add_cpts([CPT(Factor(variables=['a'], data=[0.5, 0.5]),child='a')
,CPT(Factor(variables=['a','b'], data=[0.3, 0.7, 0.4, 0.6]),child='b')
])
w = CausalWorld(bn0)
samples = w.observe(10000)
disp('two_depend', samples)
bn1 = BN(domain=Domain(), new_domain_variables={'a': [0,1], 'b':[0,1]})
bn1.add_cpts([CPT(Factor(variables=['a'], data=[0.5, 0.5]),child='a')
,CPT(Factor(variables=['b'], data=[0.3, 0.7]),child='b')
])
w = CausalWorld(bn1)
samples = w.observe(10000)
disp('two_independ', samples)
class TestParameters(unittest.TestCase):
def setUp(self):
from gPy.Variables import Domain
self.domain = Domain()
self.bnm = BN(domain=self.domain)
self.bnm.from_dnet(read_dnet('Asia.dnet'))
self.cptdict = {}
# taken directly from Netica output
self.marginals = [
Factor((('VisitAsia'),),
[0.99,0.01]),
Factor((('Tuberculosis'),),
[0.9896,0.0104]),
Factor((('Smoking'),),
[0.5,0.5]),
Factor((('Cancer'),),
[0.945,0.055]),
Factor((('TbOrCa'),),
[0.93517, 0.064828]),
Factor((('XRay'),),
[0.11029, 0.88971]),
Factor((('Bronchitis'),),
[0.55,0.45]),
Factor((('Dyspnea'),),
[0.56403,0.43597])
]
# taken directly from Netica output
self.cond_marginals = [
Factor((('VisitAsia'),),
[0.95192,0.048077]),
Factor((('Tuberculosis'),),
[0,1]),
Factor((('Smoking'),),
[0.52381,0.47619]),
#other marginals are conditional on these values
#Factor((('Cancer'),),
# [1,0]),
#Factor((('TbOrCa'),),
# [0,1]),
Factor((('XRay'),),
[0.98, 0.02]),
Factor((('Bronchitis'),),
[0.55714,0.44286]),
Factor((('Dyspnea'),),
[0.21143,0.78857])
]
for cpt in self.bnm:
self.cptdict[cpt.child()] = cpt
self.rawdata = read_csv(open('alarm_1K.dat'))
def test_str(self):
out = ''
for node in ['VisitAsia','Tuberculosis','Smoking','Cancer',
'TbOrCa','XRay','Bronchitis','Dyspnea']:
out += str(self.cptdict[node])
self.assertEqual(out,open('Asia.cpts').read())
def test_cptcheck(self):
# be nice to check that this is printed out
errmsg = """
For child: Dyspnea
For row: Bronchitis=Absent, Dyspnea=Absent, TbOrCa=True
Sum was: 0.90 (should be 1.0)
"""
def bnfromfile(filename): x= BN(domain=self.domain); x.from_dnet(read_dnet(filename)); return x
self.assertRaises(CPTError,bnfromfile,open('Asia_wrong.dnet'))
def test_multinplace(self):
factor = self.cptdict['VisitAsia'] * self.cptdict['Tuberculosis']
fid = id(factor)
factor *= self.cptdict['Bronchitis']
self.assertEqual(fid,id(factor))
factor2 = self.cptdict['VisitAsia'] * \
self.cptdict['Tuberculosis'] * self.cptdict['Bronchitis']
self.samefactor(factor,factor2)
def test_scalarprod(self):
for factor in self.bnm:
for scalar in 0,0.3,7:
tf1 = scalar * factor
tf2 = factor * scalar
self.samefactor(tf1,tf2)
for i, val in enumerate(tf1._data):
self.assertAlmostEqual(val,scalar * factor._data[i],places)
def test_instbyname(self):
dyspnea = self.cptdict['Dyspnea']
datum = dyspnea[{'Dyspnea':'Present',
'Bronchitis':'Present',
'TbOrCa':'False'}]
self.assertAlmostEqual(datum,0.8,places)
def test_getitem(self):
dyspnea = self.cptdict['Dyspnea']
datum = dyspnea['Present','Present','False']
self.assertAlmostEqual(datum,0.8,places)
def test_z(self):
factor = 1
order = ['VisitAsia','Tuberculosis','Smoking',
'Cancer','TbOrCa','XRay','Bronchitis',
'Dyspnea']
for name in order:
newfactor = factor * self.cptdict[name]
factor *= self.cptdict[name]
self.samefactor(factor,newfactor)
self.assertAlmostEqual(factor.z(),1,places)
self.assertEqual(str(factor),open('Asia.joint').read())
factor = 1
for cpt in self.cptdict.values():
factor *= cpt
self.assertAlmostEqual(factor.z(),1,places)
def test_marginals(self):
joint = 1
for cpt in self.bnm:
joint *= cpt
order = ['VisitAsia','Tuberculosis','Smoking',
'Cancer','TbOrCa','XRay','Bronchitis',
'Dyspnea']
varset = set(order)
for i, var in enumerate(order):
tmp = varset.copy()
tmp.remove(var)
self.samefactor(self.marginals[i],joint.sumout(tmp))
def test_factorprodcommute(self):
for f1 in self.bnm:
for f2 in self.bnm:
tf1 = f1 * f2
tf2 = f2 * f1
self.samefactor(tf1,tf2)
def test_restrict(self):
bnm = self.bnm.copy(copy_domain=True)
given = {'Cancer':['Absent'],'TbOrCa':['True']}
bnm.condition(given)
joint = 1
for cpt in bnm:
joint *= cpt
order = ['VisitAsia','Tuberculosis','Smoking',
'XRay','Bronchitis','Dyspnea']
varset = set(order)
varset.update(['Cancer','TbOrCa'])
for i, var in enumerate(order):
marginal = joint.sumout(varset - set([var]))
marginal /= marginal.z()
self.samefactor(self.cond_marginals[i],marginal)
def samefactor(self,tf1,tf2):
self.assertEqual(tf1.variables(),tf2.variables())
tf2data = tf2.data()
for i, val in enumerate(tf1.data()):
self.assertAlmostEqual(val,tf2data[i],places)
def bnfromfile(filename): x= BN(domain=self.domain); x.from_dnet(read_dnet(filename)); return x
self.assertRaises(CPTError,bnfromfile,open('Asia_wrong.dnet'))