def test_reduce(self):
from math import log
d = Document('http://a.com', 'hello there hello')
out = []
class mock_reduce_out(object):
def add(self, k, v):
out.append((k,v))
reduce_out = mock_reduce_out()
map_out = [('hello', 1), ('hello ', ('http://a.com', 2.0/3)),
('there', 1), ('there ', ('http://a.com', 1.0/3))]
# unlike in test_TfIdf.map, using doc_count=2
TfIdf.reduce(map_out, reduce_out, {'doc_count':2})
expected = [('hello', ('http://a.com', 2.0/3 * log(2))),
('there', ('http://a.com', 1.0/3 * log(2)))]
self.assertEqual(sorted(expected), sorted(out))
def test_map(self):
d = Document('http://a.com', 'hello there hello')
out = list(TfIdf.map(d, None))
expected = [('hello', 1), ('there', 1),
('hello ', ('http://a.com', 2.0/3)),
('there ', ('http://a.com', 1.0/3))]
self.assertEqual(sorted(expected), sorted(out))
def test_partition(self):
self.assertEquals(
TfIdf.partition('hi', 1000, None),
TfIdf.partition('hi ', 1000, None))
