def learning(self):
mln = MLN(mlnfile=self.mln_path, grammar='StandardGrammar')
db = Database(mln, dbfile=self.db_path)
for method in ('BPLL', 'BPLL_CG', 'CLL'):
print('=== LEARNING TEST:', method, '===')
learn(method=method, mln=mln, db=db, verbose=True,
multicore=False).run()
def __init__(self,
prac,
evidence=None,
db=None,
ignore_unknown_preds=False):
self.prac = prac
if evidence: pass
elif db:
evidence = db.evidence
else:
evidence = {}
Database.__init__(self,
prac.mln,
evidence=evidence,
dbfile=None,
ignore_unknown_preds=ignore_unknown_preds)
def test_learning_taxonomies():
p = '$PRACMLN_HOME/examples/taxonomies/taxonomies.pracmln'
mln = MLN(mlnfile=('%s:senses_and_roles.mln' % p), grammar='PRACGrammar')
mln.write()
dbs = Database.load(mln, dbfiles='%s:training.db' % p)
for method in ('DPLL', 'DBPLL_CG', 'DCLL'):
for multicore in (True, False):
print '=== LEARNING TEST:', method, '==='
learn(method=method, mln=mln, db=dbs, verbose=True, multicore=multicore, epreds='is_a', discr_preds=EVIDENCE_PREDS).run()
def finalgraph(self, filename=None):
finaldb = Database(self.prac.mln)
for step in self.inference_steps:
for db in step.output_dbs:
for atom, truth in list(db.evidence.items()):
if truth == 0: continue
_, predname, args = self.prac.mln.logic.parseLiteral(atom)
if predname in self.prac.roles.union(
['has_sense', 'action_core', 'achieved_by']):
finaldb << atom
# finaldb.write(sys.stdout, color=True)
g = Digraph(format='svg', engine='dot')
g.attr('node', shape='box', style='filled')
for res in finaldb.query('action_core(?w, ?a) ^ has_sense(?w, ?s)'):
actioncore = res['?a']
sense = res['?s']
predname = 'action_core'
g.node(actioncore, fillcolor='#bee280')
g.node(sense)
g.edge(actioncore, sense, label='is_a')
roles = self.prac.actioncores[actioncore].roles
for role in roles:
for res in db.query('{}(?w, {}) ^ has_sense(?w, ?s)'.format(
role, actioncore)):
sense = res['?s']
g.node(sense)
g.edge(actioncore, sense, label=role)
for res in finaldb.query('achieved_by(?a1, ?a2)'):
a1 = res['?a1']
a2 = res['?a2']
g.node(a1, fillcolor='#bee280')
g.node(a2, fillcolor='#bee280')
g.edge(a1, a2, label='achieved_by')
actioncore = a2
roles = self.prac.actionroles[actioncore].roles
for role in roles:
for res in db.query('{}(?w, {}) ^ has_sense(?w, ?s)'.format(
role, actioncore)):
sense = res['?s']
g.node(sense)
g.edge(actioncore, sense, label=role)
return render_gv(g, filename)
def test_GSMLN():
# mln = MLN(grammar='GSMLNGrammar')
# mln << 'residue(id, profile)'
# mln << 'partners(id, id)'
# f = "residue(a, $pa) v residue(b, $pb) => partners(a,b)"
# # f = "((a(x) ^ b(x)) v (c(x) ^ !(d(x) ^ e(x) ^ g(x)))) => f(x)"
# # f = "(a(x) v (b(x) ^ c(x))) => f(x)"
# f = mln.logic.grammar.parse_formula(f)
# f.print_structure()
# print(list(f.literals()))
# g = "partners(id, id)"
# g = mln.logic.grammar.parse_predicate(g)
# print(g)
# print(mln.predicates)
mln = MLN(mlnfile='smokers.mln', grammar='GSMLNGrammar')
# mln.write()
# print(mln.predicates)
dbs = Database.load(mln, dbfiles='smokers.db')
# dbs[0].write()
# print(mln.formulas[0].neural)
# print(mln.formulas[0].cnf())
# print(mln.nnformulas[0].idx)
# print(mln.domains)
# print(dbs[0].domains)
# mln.formulas[1].print_structure()
# mrf = mln.ground(dbs[0])
# grounder = DefaultGroundingFactory(mrf, simplify=False, unsatfailure=True, verbose=False, cache=0)
# for f in grounder.itergroundings():
# print(f)
# print((mrf.gndatoms))
# mln = MLN(grammar='GSMLNGrammar')
# mln << 'Cancer(&person)'
# mln << 'Friends(&person,&person)'
# mln << 'Smokes(&person)'
# f = 'Smokes($x) => Cancer($x)'
# g = 'Friends($x,$y) => (Smokes($x) <=> Smokes($y))'
# print(mln.logic.grammar.parse_formula(f))
# mln.formula(f)
# mln.formula(g)
# print(mln.predicates)
# print(mln.formulas)
# mln.formulas[0].print_structure()
# print(mln.domains)
# print(mln.formulas[0].cnf())
# this uses the method from base.py
learned_mln = mln.learn(databases=dbs, method=GSMLN_L, verbose=True)
def test_inference_smokers():
p = os.path.join(locs.examples, 'smokers', 'smokers')
# mln = MLN(mlnfile=('%s:wts.pybpll.smoking-train-smoking.mln' % p),
# grammar='StandardGrammar')
# db = Database(mln, dbfile='%s:smoking-test-smaller.db' % p)
print(p)
mln = MLN(mlnfile=('%s.mln' % p), grammar='StandardGrammar')
db = Database(mln, dbfile='%s.db' % p)
for method in ('GibbsSampler', ):
print('=== INFERENCE TEST:', method, '===')
query(queries='Cancer,Smokes,Friends',
method=method,
mln=mln,
db=db,
verbose=True,
multicore=False).run()
def test_learning_smokers(arg='.', disp=False):
pth = os.path.join(arg, 'smoking.mln')
mln = MLN(mlnfile=pth, grammar='StandardGrammar')
pth = os.path.join(arg, 'smoking-train.db')
db = Database(mln, dbfile=pth)
for method in ('BPLL', 'BPLL_CG', 'CLL'):
for multicore in (True, False):
print('=== LEARNING TEST:', method, '===')
if disp:
query(queries='Cancer,Smokes,Friends',
method=method,
mln=mln,
db=db,
verbose=disp,
multicore=multicore).run().write()
else:
learn(method=method,
mln=mln,
db=db,
verbose=disp,
multicore=multicore).run()
def inference(self, inference_query):
mln = MLN(mlnfile=self.mln_path, grammar='StandardGrammar')
db = Database(mln, dbfile=self.db_path)
for method in [
'EnumerationAsk'
# 'MC-SAT',
# 'WCSPInference',
# 'GibbsSampler'
]:
print('=== INFERENCE TEST:', method, '===')
result = query(queries=inference_query,
method=method,
mln=mln,
db=db,
verbose=True,
multicore=False).run()
print(result)
if result:
return result
else:
return []
def test_inference_smokers(arg='.', disp=False):
pth = os.path.join(arg, 'wts.pybpll.smoking-train-smoking.mln')
mln = MLN(mlnfile=pth, grammar='StandardGrammar')
pth = os.path.join(arg, 'smoking-test-smaller.db')
db = Database(mln, dbfile=pth)
for method in ('EnumerationAsk', 'MC-SAT', 'WCSPInference',
'GibbsSampler'):
for multicore in (False, True):
print('=== INFERENCE TEST:', method, '===')
if disp:
query(queries='Cancer,Smokes,Friends',
method=method,
mln=mln,
db=db,
verbose=disp,
multicore=multicore).run().write()
else:
query(queries='Cancer,Smokes,Friends',
method=method,
mln=mln,
db=db,
verbose=disp,
multicore=multicore).run()
