def run():
# ds = AnnotatedDataset.parse_abbadingoformat("../../testdata/RPNIAndEDSMDataSample.txt")
ds = AnnotatedDataset.parse_custom("testdata/language.txt")
# print ds
# ds = AnnotatedDataset.parse_abbadingoformat("../testdata/train170.txt")
# labelspath = "../testdata/lol.labels"
# datapath = "../testdata/lol.data"
# ds = AnnotatedDataset.parse_sentence(labelspath, datapath)
dfa = build_gold_DFA(ds)
dfa.write_png()
for sample in ds:
assert (dfa.membership(sample[0]) == sample[1])
print dfa
def run():
# labelspath = "../testdata/cms.01.labels"
# datapath = "../testdata/cms.01.data"
# labelspath = "../testdata/fun.labels"
# datapath = "../testdata/fun.data"
# ds = AnnotatedDataset.parse_sentence(labelspath, datapath)
# ds = AnnotatedDataset.parse_abbadingoformat("../testdata/train170.txt")
# ds = AnnotatedDataset.parse_ewsformat_bin(labelspath, datapath)
ds = AnnotatedDataset.parse_abbadingoformat(
"testdata/RPNIAndEDSMDataSample.txt")
pta = build_PTA(ds)
# pta3 = pta.to_reverse_NFA().to_DFA().to_reverse_NFA().to_DFA()
# pta.set_error_state()
# print pta
# print pta3 #89 states without error state --> 90
# pta2 = pta.minimize()
# print pta2
# print pta == pta2
# eqdfa = pta.symmetric_difference(pta2)
# print eqdfa
# for i in eqdfa.generate():
# print i, pta.is_member(i), pta2.is_member(i)
# pta.minimize()
# if not pta == pta2:
# print "not equal?"
# strng = (pta ^ pta2).shortest_example()
# print strng, pta.parse(strng), pta2.parse(strng)
# else:
# print "equal"
# print
# for sample in ds:
# p = pta2.parse(sample[0])
# if p[0] != sample[1]:
# print p, sample
# print pta == pta3
# print pta2 == pta3
# for sample in ds:
# assert(pta.is_member(sample[0]) == sample[1])
# dfa = pta.minimize().rename().del_dead_states()
dfa = pta
dfa.write_png()
def run():
# labelspath = "../testdata/fun.labels"
# datapath = "../testdata/fun.data"
# ds = AnnotatedDataset.parse_sentence(labelspath, datapath)
ds = AnnotatedDataset.parse_abbadingoformat("../../testdata/train171.txt")
# ds = AnnotatedDataset.parse_abbadingoformat("../testdata/RPNIAndEDSMDataSample.txt")
# labelspath = "../testdata/cms.01.labels"
# datapath = "../testdata/cms.01.data"
# ds = AnnotatedDataset.parse_ewsformat_bin(labelspath, datapath)
# labelspath = "../testdata/lol.labels"
# datapath = "../testdata/lol.data"
# ds = AnnotatedDataset.parse_sentence(labelspath, datapath)
dfa = build_rpni_DFA(ds)
# dfa.write_png()
print dfa.states, dfa.accept, dfa.reject
for sample in ds:
# res, state = dfa.parse_final(sample[0])
res = dfa.membership(sample[0])
if res != sample[1]:
print sample, res
print dfa
def run():
# datapath = "../testdata/fun.data"
# ds = AnnotatedDataset.parse_sentence(labelspath, datapath)
# ds = Dataset.parse_sentence(datapath)
# ds = AnnotatedDataset.parse_ewsformat(labelspath, datapath)
ds = AnnotatedDataset.parse_abbadingoformat("testdata/train173.txt")
# ds = AnnotatedDataset.parse_abbadingoformat("../../testdata/RPNIAndEDSMDataSample.txt")
oracle = MinAdequateTeacher.from_AnnotatedDataset(ds)
pta = build_PTA(ds)
dfa = build_lstar_DFA(oracle)
print dfa
print pta == dfa
# eqdfa = pta.symmetric_difference(pta2)
# print eqdfa
# for i in eqdfa.generate():
# print i, pta.is_member(i), pta2.is_member(i)
# pta.minimize()
# print pta == pta2
# print pta == pta3
# print pta2 == pta3
# for sample in ds:
# assert(pta.is_member(sample[0]) == sample[1])
dfa.del_dead_states()
dfa.write_png()
def setUp(self):
self.raw = [("bb", 1), ("abb", 1), ("bba", 1), ("bbb", 1), ("a", 0),
("b", 0), ("aa", 0), ("bab", 0)]
self.ds = AnnotatedDataset.from_tuples(self.raw)