def test_transition_model(self):
# Links exist
corpus = {
"1.html": {"2.html", "3.html"},
"2.html": {"3.html"},
"3.html": {"2.html"},
}
page = "1.html"
damping_factor = 0.85
result = {"1.html": 0.05, "2.html": 0.475, "3.html": 0.475}
for page, rank in pr.transition_model(corpus, page,
damping_factor).items():
self.assertAlmostEqual(result[page], rank)
# No links exist
corpus = {
"1.html": {},
"2.html": {"3.html"},
"3.html": {"2.html"},
}
page = "1.html"
damping_factor = 0.85
result = {"1.html": 0.333, "2.html": 0.333, "3.html": 0.333}
for page, rank in pr.transition_model(corpus, page,
damping_factor).items():
self.assertAlmostEqual(result[page], rank, 3)
def testTransitionModel(self):
trans_model_0 = pr.transition_model(self.corpus_0, "1.html", DAMPING)
self.assertEqual(trans_model_0, {"1.html": 0.05, "2.html": 0.475, "3.html": 0.475})
trans_model_1 = pr.transition_model(self.corpus_1, "1.html", DAMPING)
self.assertEqual(trans_model_1, {"1.html": 0.3333333333333333, "2.html": 0.3333333333333333, "3.html": 0.3333333333333333})
trans_model_2 = pr.transition_model(self.corpus_2, "2.html", DAMPING)
self.assertEqual(trans_model_2, {"2.html": 0.500, "3.html": 0.500})
def test_transition_model_no_links(self):
model_1 = transition_model(
{"1.html": {}, "2.html": {"3.html"}, "3.html": {"2.html"}},
"1.html",
0.85
)
model_2 = {"1.html": 0.33, "2.html": 0.33, "3.html": 0.33}
for link in model_1:
if link in model_2:
self.assertAlmostEqual(model_1[link], model_2[link], places=2)
def test_transition_model_example(self):
corpus = {
"1.html": {"2.html", "3.html"},
"2.html": {"3.html"},
"3.html": {"2.html"}
}
page = "1.html"
damping_factor = 0.85
expected = {"1.html": 0.05, "2.html": 0.475, "3.html": 0.475}
result = transition_model(corpus, page, damping_factor)
self.assertEqual(result, expected)
def test_transition_model_no_links(self):
corpus = {
"1.html": {"2.html", "3.html"},
"2.html": {"3.html"},
"3.html": {"2.html"},
"4.html": {}
}
page = "4.html"
damping_factor = 0.85
expected = {
"1.html": 0.25,
"2.html": 0.25,
"3.html": 0.25,
"4.html": 0.25
}
result = transition_model(corpus, page, damping_factor)
self.assertEqual(result, expected)
from pagerank import transition_model
transition_model(transition_model('corpus0', '1.html', 0.85))
def test_transition_model():
assert pagerank.transition_model(CRAWL0, "1.html") == PAGE_PROBS_1
assert pagerank.transition_model(CRAWL0, "2.html") == PAGE_PROBS_2
assert pagerank.transition_model(CRAWL0, "3.html") == PAGE_PROBS_3
assert pagerank.transition_model(CRAWL0, "4.html") == PAGE_PROBS_4
from pagerank import transition_model, crawl
if len(sys.argv) != 2:
sys.exit("Usage: python pagerank.py corpus")
corpus = crawl(sys.argv[1])
print(corpus)
trans = transition_model(corpus, "1.html", 0.85)
print(trans)
from pagerank import transition_model, sample_pagerank, iterate_pagerank
corpus = {
"1.html": {"2.html", "3.html"},
"2.html": {"3.html"},
"3.html": {"2.html"}
}
page = "1.html"
damp = 0.85
transition_model(corpus, page, damp)
print(sample_pagerank(corpus, 0.85, 5000))
print(iterate_pagerank(corpus, 0.85))