def test_similar_row_from_datum_and_rate(self):
filter_warning()
recommender = Recommender.run(Config())
loader = StubLoader()
schema = Schema({'v': Schema.ID})
dataset = Dataset(loader, schema)
for (idx, row_id,
result) in recommender.similar_row_from_datum_and_rate(dataset):
self.assertEqual(0, len(result))
# rate must be in (0, 1].
def func():
for _ in recommender.similar_row_from_datum_and_rate(dataset,
rate=0.0):
pass
self.assertRaises(ValueError, lambda: func())
def func():
for _ in recommender.similar_row_from_datum_and_rate(dataset,
rate=1.01):
pass
self.assertRaises(ValueError, lambda: func())
schema = Schema({'v': Schema.NUMBER})
dataset = Dataset(loader, schema)
for (idx, row_id,
result) in recommender.similar_row_from_datum_and_rate(dataset):
self.assertEqual(None, row_id) # there is no id in column_table
self.assertEqual(
0, len(result)) # there is no similar row in column_table
recommender.stop()
def test_similar_row_from_id_and_score(self):
filter_warning()
recommender = Recommender.run(Config())
loader = StubLoader()
# dataset must have id when execute `similar_row_from_id_and_score`
schema = Schema({'v': Schema.NUMBER})
dataset = Dataset(loader, schema)
def func():
for _ in recommender.similar_row_from_id_and_score(dataset):
pass
self.assertRaises(RuntimeError, lambda: func())
schema = Schema({'v': Schema.ID})
dataset = Dataset(loader, schema)
for (idx, row_id,
result) in recommender.similar_row_from_id_and_score(dataset):
self.assertEqual(str(idx + 1),
row_id) # there is no id in column_table
self.assertEqual(
0, len(result)) # there is no similar row in column_table
recommender.stop()
def test_method_param(self):
self.assertTrue('parameter' not in Config(method='inverted_index'))
self.assertTrue('hash_num' in Config(method='minhash')['parameter'])
self.assertTrue('hash_num' in Config(method='lsh')['parameter'])
self.assertTrue('threads' in Config(method='lsh')['parameter'])
self.assertTrue('method' in Config(
method='nearest_neighbor_recommender')['parameter'])
self.assertTrue('parameter' in Config(
method='nearest_neighbor_recommender')['parameter'])
self.assertTrue('threads' in Config(
method='nearest_neighbor_recommender')['parameter']['parameter'])
self.assertTrue('hash_num' in Config(
method='nearest_neighbor_recommender')['parameter']['parameter'])
def test_update_row(self):
recommender = Recommender.run(Config())
loader = StubLoader()
# dataset must have id when execute `update_row`
schema = Schema({'v': Schema.NUMBER})
dataset = Dataset(loader, schema)
def func():
for _ in recommender.update_row(dataset):
pass
self.assertRaises(RuntimeError, lambda: func())
schema = Schema({'v': Schema.ID})
dataset = Dataset(loader, schema)
for (idx, row_id, result) in recommender.update_row(dataset):
self.assertEqual(result, True)
recommender.stop()
def test_similar_row_from_datum(self):
filter_warning()
recommender = Recommender.run(Config())
loader = StubLoader()
schema = Schema({'v': Schema.ID})
dataset = Dataset(loader, schema)
for (idx, row_id,
result) in recommender.similar_row_from_datum(dataset):
self.assertEqual(0, len(result))
schema = Schema({'v': Schema.NUMBER})
dataset = Dataset(loader, schema)
for (idx, row_id,
result) in recommender.similar_row_from_datum(dataset):
self.assertEqual(None, row_id) # there is no id in column_table
self.assertEqual(
0, len(result)) # there is no similar row in column_table
recommender.stop()
def test_complete_row_from_datum(self):
filter_warning()
recommender = Recommender.run(Config())
loader = StubLoader()
schema = Schema({'v': Schema.ID})
dataset = Dataset(loader, schema)
for (idx, row_id, d) in recommender.complete_row_from_datum(dataset):
self.assertEqual(0, len(d.string_values))
self.assertEqual(0, len(d.num_values))
self.assertEqual(0, len(d.binary_values))
schema = Schema({'v': Schema.NUMBER})
dataset = Dataset(loader, schema)
for (idx, row_id, d) in recommender.complete_row_from_datum(dataset):
self.assertEqual(None, row_id) # there is no id in column_table.
self.assertEqual(0, len(d.string_values))
self.assertEqual(0, len(d.num_values))
self.assertEqual(0, len(d.binary_values))
recommender.stop()
def test_clear_row(self):
recommender = Recommender.run(Config())
loader = StubLoader()
# dataset must have id when execute `clear_row`.
schema = Schema({'v': Schema.NUMBER})
dataset = Dataset(loader, schema)
def func():
for _ in recommender.clear_row(dataset):
pass
self.assertRaises(RuntimeError, lambda: func())
schema = Schema({'v': Schema.ID})
dataset = Dataset(loader, schema)
# expect to get False when table is empty.
for (idx, row_id, result) in recommender.clear_row(dataset):
self.assertEqual(result, True)
recommender.stop()
def test_complete_row_from_id(self):
filter_warning()
recommender = Recommender.run(Config())
loader = StubLoader()
# dataset must have id when execute `complete_row_from_id`
schema = Schema({'v': Schema.NUMBER})
dataset = Dataset(loader, schema)
def func():
for _ in recommender.complete_row_from_id(dataset):
pass
self.assertRaises(RuntimeError, lambda: func())
schema = Schema({'v': Schema.ID})
dataset = Dataset(loader, schema)
for (idx, row_id, d) in recommender.complete_row_from_id(dataset):
self.assertEqual(0, len(d.string_values))
self.assertEqual(0, len(d.num_values))
self.assertEqual(0, len(d.binary_values))
recommender.stop()
def test_embedded(self):
recommender = Recommender.run(Config(), embedded=True)
def test_default(self):
config = Config.default()
self.assertEqual('lsh', config['method'])
def test_methods(self):
config = Config()
self.assertTrue(isinstance(config.methods(), list))
def test_simple(self):
config = Config()
self.assertEqual('lsh', config['method'])
def test_default(self):
config = Config.default()
self.assertEqual('lsh', config['method'])
def test_methods(self): config = Config() self.assertTrue(isinstance(config.methods(), list))
from jubakit.loader.csv import CSVLoader
# Load a CSV file.
loader = CSVLoader('npb.csv')
# Define a Schema that defines types for each columns of the CSV file.
schema = Schema({
'name': Schema.ID,
'team': Schema.STRING,
}, Schema.NUMBER)
# Create a Dataset.
dataset = Dataset(loader, schema)
# Create an Recommender Service.
cfg = Config(method='lsh')
recommender = Recommender.run(cfg)
# Update the Recommender model.
for (idx, row_id, success) in recommender.update_row(dataset):
pass
# Calculate the similarity in recommender model from row-id and display top-2 similar items.
print('{0}\n recommend similar players from row-id \n{1}'.format(
'-' * 60, '-' * 60))
for (idx, row_id, result) in recommender.similar_row_from_id(dataset, size=3):
if idx % 10 == 0:
print(
'player {0} is similar to : {1} (score:{2:.3f}), {3} (score:{4:.3f})'
.format(result[0].id, result[1].id, result[1].score, result[2].id,
result[2].score))