def main():
args = parse_options()
client = Classifier('127.0.0.1', args.port, 'test', 0)
d = Datum()
# Learn same data
rand = random.randint(0, 1)
d.add_number('key', 1.0 if rand else 2.0)
print client.classify([d])
print client.get_labels()
def main():
args = parse_options()
client = Classifier('127.0.0.1', args.port, 'test', 0)
d = Datum()
# Learn same data
rand = random.randint(0, 1)
d.add_number('key', 1.0 if rand else 2.0)
print client.classify([d])
print client.get_labels()
def get_classify_data(usr):
user = usr['display_name']
print user
options, remainder = parse_args()
classifier = Classifier(options.server_ip,options.server_port, options.name, 10.0)
#train
data_reader = select_Interrest_Blog(user)
for row in data_reader:
label = row['CATEGORY']
dat = row['TITLE']
datum = Datum({"message": dat})
classifier.train([LabeledDatum(label, datum)])
url_list = []
url_list = get_rss_data_from_catlist(usr,['social','fun','entertainment','game'])
for data in url_list:
title = data["title"]
datum = Datum({"message": title})
classifier.train([LabeledDatum('no', datum)])
# print classifier.get_status()
# print classifier.save("tutorial")
# print classifier.load("tutorial")
# print classifier.get_config()
url_list = []
ret1 = []
ret2 = []
url_list = get_rss_data_from_catlist(usr,['it','popular','life','knowledge'])
for data in url_list:
title = data["title"]
datum = Datum({"message": title})
ans = classifier.classify([datum])
if ans != None:
estm = get_most_likely(ans[0])
if estm[0] == 'yes':
ret1.append(data)
else:
ret2.append(data)
print ret1
print ""
print ret2
return ret1,ret2
datum = Datum({"message": dat})
classifier.train([LabeledDatum(label, datum)])
print classifier.get_status()
print classifier.save("tutorial")
print classifier.load("tutorial")
print classifier.get_config()
count_ok = 0
count_ng = 0
for line in open('test.dat'):
label, file = line[:-1].split(',')
dat = open(file).read()
datum = Datum({"message": dat})
ans = classifier.classify([datum])
if ans != None:
estm = get_most_likely(ans[0])
if (label == estm[0]):
result = "OK"
count_ok += 1
else:
result = "NG"
count_ng += 1
print result + "," + label + ", " + estm[0] + ", " + str(estm[1])
print "==================="
print "OK: {0}".format(count_ok)
print "NG: {0}".format(count_ng)
class ClassifierTest(unittest.TestCase):
def setUp(self):
self.config = {
"method": "AROW",
"converter": {
"string_filter_types": {},
"string_filter_rules": [],
"num_filter_types": {},
"num_filter_rules": [],
"string_types": {},
"string_rules": [{"key": "*", "type": "str", "sample_weight": "bin", "global_weight": "bin"}],
"num_types": {},
"num_rules": [{"key": "*", "type": "num"}],
},
"parameter": {"regularization_weight": 1.001},
}
TestUtil.write_file("config_classifier.json", json.dumps(self.config))
self.srv = TestUtil.fork_process("classifier", port, "config_classifier.json")
try:
self.cli = Classifier(host, port, "name")
except:
TestUtil.kill_process(self.srv)
raise
def tearDown(self):
if self.cli:
self.cli.get_client().close()
TestUtil.kill_process(self.srv)
def test_get_client(self):
self.assertTrue(isinstance(self.cli.get_client(), msgpackrpc.client.Client))
def test_get_config(self):
config = self.cli.get_config()
self.assertEqual(json.dumps(json.loads(config), sort_keys=True), json.dumps(self.config, sort_keys=True))
def test_train(self):
d = Datum({"skey1": "val1", "skey2": "val2", "nkey1": 1.0, "nkey2": 2.0})
data = [["label", d]]
self.assertEqual(self.cli.train(data), 1)
def test_classify(self):
d = Datum({"skey1": "val1", "skey2": "val2", "nkey1": 1.0, "nkey2": 2.0})
data = [d]
result = self.cli.classify(data)
def test_set_label(self):
self.assertEqual(self.cli.set_label("label"), True)
def test_get_labels(self):
self.cli.set_label("label")
self.assertEqual(self.cli.get_labels(), {"label": 0})
def test_delete_label(self):
self.cli.set_label("label")
self.assertEqual(self.cli.delete_label("label"), True)
def test_save(self):
self.assertEqual(len(self.cli.save("classifier.save_test.model")), 1)
def test_load(self):
model_name = "classifier.load_test.model"
self.cli.save(model_name)
self.assertEqual(self.cli.load(model_name), True)
def test_get_status(self):
self.cli.get_status()
def test_str(self):
self.assertEqual("estimate_result{label: label, score: 1.0}", str(EstimateResult("label", 1.0)))
def main():
client = Classifier("127.0.0.1", port, "sleeping", timeout)
client.classify([])
datum = Datum({"message": dat})
classifier.train([LabeledDatum(label, datum)])
print classifier.get_status()
print classifier.save("tutorial")
print classifier.load("tutorial")
print classifier.get_config()
count_ok = 0
count_ng = 0
for line in open('test.dat'):
label, file = line[:-1].split(',')
dat = open(file).read()
datum = Datum({"message": dat})
ans = classifier.classify([datum])
if ans != None:
estm = get_most_likely(ans[0])
if (label == estm[0]):
result = "OK"
count_ok += 1
else:
result = "NG"
count_ng += 1
print result + "," + label + ", " + estm[0] + ", " + str(estm[1])
print "==================="
print "OK: {0}".format(count_ok)
print "NG: {0}".format(count_ng)
class ClassifierTest(unittest.TestCase):
def setUp(self):
self.config = {
"method": "AROW",
"converter": {
"string_filter_types": {},
"string_filter_rules": [],
"num_filter_types": {},
"num_filter_rules": [],
"string_types": {},
"string_rules": [{
"key": "*",
"type": "str",
"sample_weight": "bin",
"global_weight": "bin"
}],
"num_types": {},
"num_rules": [{
"key": "*",
"type": "num"
}]
},
"parameter": {
"regularization_weight": 1.001
}
}
TestUtil.write_file('config_classifier.json', json.dumps(self.config))
self.srv = TestUtil.fork_process('classifier', port,
'config_classifier.json')
try:
self.cli = Classifier(host, port, "name")
except:
TestUtil.kill_process(self.srv)
raise
def tearDown(self):
if self.cli:
self.cli.get_client().close()
TestUtil.kill_process(self.srv)
def test_get_client(self):
self.assertTrue(
isinstance(self.cli.get_client(), msgpackrpc.client.Client))
def test_get_config(self):
config = self.cli.get_config()
self.assertEqual(json.dumps(json.loads(config), sort_keys=True),
json.dumps(self.config, sort_keys=True))
def test_train(self):
d = Datum({
"skey1": "val1",
"skey2": "val2",
"nkey1": 1.0,
"nkey2": 2.0
})
data = [["label", d]]
self.assertEqual(self.cli.train(data), 1)
def test_classify(self):
d = Datum({
"skey1": "val1",
"skey2": "val2",
"nkey1": 1.0,
"nkey2": 2.0
})
data = [d]
result = self.cli.classify(data)
def test_set_label(self):
self.assertEqual(self.cli.set_label("label"), True)
def test_get_labels(self):
self.cli.set_label("label")
self.assertEqual(self.cli.get_labels(), {"label": 0})
def test_delete_label(self):
self.cli.set_label("label")
self.assertEqual(self.cli.delete_label("label"), True)
def test_save(self):
self.assertEqual(len(self.cli.save("classifier.save_test.model")), 1)
def test_load(self):
model_name = "classifier.load_test.model"
self.cli.save(model_name)
self.assertEqual(self.cli.load(model_name), True)
def test_get_status(self):
self.cli.get_status()
def test_str(self):
self.assertEqual("estimate_result{label: label, score: 1.0}",
str(EstimateResult("label", 1.0)))
