def testJoin():
iss = cItemsets.cItemsets(3)
a = cAssociationRules()
u1 = urlparse.urlparse('http://eins')
u2 = urlparse.urlparse('http://zwei')
u3 = urlparse.urlparse('http://drei')
u4 = urlparse.urlparse('http://vier')
u5 = urlparse.urlparse('http://fuenf')
i1 = cItemset.cItemset()
i1.SetUrls([u1,u2,u3])
i2 = cItemset.cItemset()
i2.SetUrls([u1,u2,u4])
i3 = cItemset.cItemset()
i3.SetUrls([u1,u3,u4])
i4 = cItemset.cItemset()
i4.SetUrls([u1,u3,u5])
i5 = cItemset.cItemset()
i5.SetUrls([u2,u3,u4])
iss.AddItemset(i1)
iss.AddItemset(i2)
iss.AddItemset(i3)
iss.AddItemset(i4)
iss.AddItemset(i5)
a.Join(iss).Print()
def test():
"""Built-in test method for this class."""
items = cItemsets(2)
items.OpenFile('/tmp/foo.xml')
urltuple = urlparse.urlparse('http://slashdot.org')
urltuple2 = urlparse.urlparse('http://harth.org')
itemset1 = cItemset.cItemset()
itemset2 = cItemset.cItemset()
itemset1.SetUrls([urltuple, urltuple])
itemset1.SetCount(4711)
itemset1.SetUrls([urltuple2, urltuple2])
itemset1.SetCount(4712)
items.AddItemset(itemset1)
items.AddItemset(itemset2)
items.CloseFile()
items2 = cItemsets(2)
items2.OpenFile('/tmp/foo.xml')
items2.Print()
#XXXprint 'now prune...'
#items2.Prune(2)
items2.Print()
items2.CloseFile()
def SetElements(self, lEls):
"""Read elements into internal representation.
lEls -- elements
<itemset count="74">
<url>http://slashdot.org</url>
...
</itemset>
"""
for el in lEls:
itemset = cItemset.cItemset()
itemset.SetElement(el)
self.lData.append(itemset)
def ComputeCandidateOneItemsets(self, lSessions):
"""Computes candidate one itemsets from session
lSessions -- list of sessions
return -- candidate one itemsets
"""
oneitemsets = cItemsets.cItemsets(1)
for session in lSessions:
for click in session.GetClicks():
iset = cItemset.cItemset()
iset.SetUrls([click.GetUrl(),])
oneitemsets.AddItemset(iset)
return oneitemsets
def Join(self, itemsets):
"""Join itemsets.
itemsets -- Lk-1 large k-1 itemsets
return -- Ck candidate k itemsets
"""
k = itemsets.GetSize() + 1
candidates = cItemsets.cItemsets(k)
lJoined = []
for item_i in itemsets.GetList():
for item_j in itemsets.GetList():
list_i = item_i.GetUrls()
list_j = item_j.GetUrls()
# use dict to eliminate double occurences
dict = {}
for url in list_i:
dict[url] = 1
for url in list_j:
dict[url] = 1
if len(dict) == k:
if dict not in lJoined:
lJoined.append(dict)
for dict in lJoined:
i = cItemset.cItemset()
for url in dict.keys():
i.AddUrl(url)
candidates.AddItemset(i)
if len(candidates.lData) == 0:
return None
else:
return candidates
