class TestComputation(unittest.TestCase):
debug = True
def setUp(self):
#super(Indexing, self).__init__(training_id, db_location, local_execution = True)
self.wfb = WorkFlowBase()
self.wfb.setUp()
def tearDown(self):
self.wfb.tearDown()
def importTrainingData(self, backends):
if not self.debug:
self.getResults()
n_maxrows = 10 #amount of items fetched for training
result = r.fetch_row(maxrows=n_maxrows) # fetch N row maximum
else:
result = self.wfb.mockMysqlResults_domain418()
result2 = self.wfb.mockMysqlResults_domain800()
constraints = {'domainid': 418}
for result_item in result:
json_string = result_item[3]
sM = SaveMessage()
sM.save(json_string, async=False, api='contest', backends=backends, constraints=constraints)
constraints = {'domainid': 800}
for result_item in result2:
json_string = result_item[3]
sM = SaveMessage()
sM.save(json_string, async=False, api='contest', backends=backends, constraints=constraints)
for result_item in result:
json_string = result_item[3]
sM = SaveMessage()
sM.save(json_string, async=False, api='contest', backends=backends, constraints={})
for result_item in result2:
json_string = result_item[3]
sM = SaveMessage()
sM.save(json_string, async=False, api='contest', backends=backends, constraints={})
def test_train_recommender_with_impressions_only_contest(self):
"""
todo this is an intermediate unit Test and it might be changed later on when the framework evolves
"""
# backends = [config_global.SAVE_HADOOP_SINK, config_global.SAVE_RANDOM_RECOMMENDER]
backends = [config_global.SAVE_RANDOM_RECOMMENDER]
self.debug = True
# todo something do something with the data
self.importTrainingData(backends)
message_418 = "{\"msg\":\"impression\",\"id\":1300277,\"client\":{\"id\":1},\"domain\":{\"id\":\"418\"},\"item\":{\"id\":\"123\",\"title\":\"TEXT1\",\"url\":\"http:\/\/www.ksta.de\/html\/fotolines\/1321650273296\/rahmen.shtml?1\",\"created\":1321957664,\"text\":\"Talkshow-Mode\u00adra\u00adtorin Wendy Williams bei den 39. inter\u00adna\u00adtio\u00adnalen Emmy-Awards. (Bild: AFP)\",\"img\":null,\"recommendable\":true},\"config\":{\"timeout\":null,\"recommend\":true,\"limit\":1,\"team\":{\"id\":\"22\"}},\"version\":\"1.0\"}"
gR = OnsiteRecommender(message_418, async=False, api='contest', backends=backends)
recList = gR.recommend()
print recList
self.assertGreater(len(recList), 0)
message_800 = "{\"msg\":\"impression\",\"id\":1300280,\"client\":{\"id\":1},\"domain\":{\"id\":\"800\"},\"item\":{\"id\":\"999\",\"title\":\"TEXT1\",\"url\":\"http:\/\/www.ksta.de\/html\/fotolines\/1321650273296\/rahmen.shtml?1\",\"created\":1321957664,\"text\":\"Talkshow-Mode\u00adra\u00adtorin Wendy Williams bei den 39. inter\u00adna\u00adtio\u00adnalen Emmy-Awards. (Bild: AFP)\",\"img\":null,\"recommendable\":true},\"config\":{\"timeout\":null,\"recommend\":true,\"limit\":1,\"team\":{\"id\":\"22\"}},\"version\":\"1.0\"}"
gR = OnsiteRecommender(message_800, async=False, api='contest', backends=backends)
recList = gR.recommend()
print recList
self.assertGreater(len(recList), 0)
# todo sleep for a short while
# todo get a recommendation now
def test_train_recommender_with_impressions_only_contest_Crosssite(self):
"""
todo this is an intermediate unit Test and it might be changed later on when the framework evolves
"""
# backends = [config_global.SAVE_HADOOP_SINK, config_global.SAVE_RANDOM_RECOMMENDER]
backends = [config_global.SAVE_RANDOM_RECOMMENDER]
self.debug = True
# todo something do something with the data
self.importTrainingData(backends)
message_418 = "{\"msg\":\"impression\",\"id\":1300277,\"client\":{\"id\":1},\"domain\":{\"id\":\"418\"},\"item\":{\"id\":\"123\",\"title\":\"TEXT1\",\"url\":\"http:\/\/www.ksta.de\/html\/fotolines\/1321650273296\/rahmen.shtml?1\",\"created\":1321957664,\"text\":\"Talkshow-Mode\u00adra\u00adtorin Wendy Williams bei den 39. inter\u00adna\u00adtio\u00adnalen Emmy-Awards. (Bild: AFP)\",\"img\":null,\"recommendable\":true},\"config\":{\"timeout\":null,\"recommend\":true,\"limit\":1,\"team\":{\"id\":\"22\"}},\"version\":\"1.0\"}"
gR = CrosssiteRecommender(message_418, async=False, api='contest', backends=backends)
recList = gR.recommend()
print recList
self.assertGreater(len(recList), 0)
def setUp(self):
#super(Indexing, self).__init__(training_id, db_location, local_execution = True)
self.wfb = WorkFlowBase()
self.wfb.setUp()