def setUp(self):
self.database = TransactionDatabase(1)
self.database.add(Transaction(1, [1, 2, 3], 0))
self.database.add(Transaction(2, [2, 3, 4], 1))
self.database.add(Transaction(3, [1], 0))
self.database.add(Transaction(4, [2], 1))
self.database.add(Transaction(5, [2, 3, 5], 1))
def setUp(self):
self.database = TransactionDatabase(1)
self.database.add(Transaction(1, [1,2,3], 0))
self.database.add(Transaction(2, [2,3,4], 1))
self.database.add(Transaction(3, [1], 0))
self.database.add(Transaction(4, [2], 1))
self.database.add(Transaction(5, [2,3, 5], 1))
class TransactionDatabaseTestCase(unittest.TestCase):
def setUp(self):
self.database = TransactionDatabase(1)
self.database.add(Transaction(1, [1,2,3], 0))
self.database.add(Transaction(2, [2,3,4], 1))
self.database.add(Transaction(3, [1], 0))
self.database.add(Transaction(4, [2], 1))
self.database.add(Transaction(5, [2,3, 5], 1))
def test_len(self):
self.assertEquals(len(self.database), 5)
def test_buildConditionalDatabase(self):
cond_db = self.database.buildConditionalDatabase([1])
self.assertEquals(len(cond_db), 2)
def test_transactionListFromPattern(self):
tlist = self.database.transactionListFromPattern([1])
self.assertEquals(tlist, [1, 3])
def test_labelSupport(self):
self.assertEquals(self.database.labelSupport(), 0.6)
def test_labelAndPatternSupport(self):
# All matched
self.assertEquals(self.database.labelAndPatternSupport([2]), 0.6)
# One matched
self.assertEquals(self.database.labelAndPatternSupport([5]), 0.2)
def test_patternSupport(self):
self.assertEquals(self.database.patternSupport([2,3,5]), 0.2)
def test_removeTransactions(self):
self.database.removeTransactions([3, 5])
self.assertEquals(len(self.database), 3)
def test_cleanAndPrune(self):
# Not needed anymore? Add tests later
pass
if __name__ == "__main__":
usage = "usage: %prog [options] filename"
parser = OptionParser(usage)
parser.add_option("-d", "--debug",
action="store_true", dest="debug")
(options, args) = parser.parse_args()
if len(args) != 3:
parser.error("You must give the filename, label support symbol, and minimum support.")
print "Mining from file %s..." % args[0]
print "Using label support symbol %s..." % args[1]
print "Using support of %s..." % args[2]
database = TransactionDatabase.loadFromFile(args[0],args[1],int(float(args[2])))
miner = DDPMine(debug=False)
miner.mine(database,int(float(args[2])))
"""
database = TransactionDatabase.loadFromFile("test.csv")
database.cleanAndPrune(2)
print "Cleaned database:"
print database
print "\nItems in FP tree and corresponding nodes:"
tree = FPTree()
for t in database:
tree.add(t)
print str(tree)
class TransactionDatabaseTestCase(unittest.TestCase):
def setUp(self):
self.database = TransactionDatabase(1)
self.database.add(Transaction(1, [1, 2, 3], 0))
self.database.add(Transaction(2, [2, 3, 4], 1))
self.database.add(Transaction(3, [1], 0))
self.database.add(Transaction(4, [2], 1))
self.database.add(Transaction(5, [2, 3, 5], 1))
def test_len(self):
self.assertEquals(len(self.database), 5)
def test_buildConditionalDatabase(self):
cond_db = self.database.buildConditionalDatabase([1])
self.assertEquals(len(cond_db), 2)
def test_transactionListFromPattern(self):
tlist = self.database.transactionListFromPattern([1])
self.assertEquals(tlist, [1, 3])
def test_labelSupport(self):
self.assertEquals(self.database.labelSupport(), 0.6)
def test_labelAndPatternSupport(self):
# All matched
self.assertEquals(self.database.labelAndPatternSupport([2]), 0.6)
# One matched
self.assertEquals(self.database.labelAndPatternSupport([5]), 0.2)
def test_patternSupport(self):
self.assertEquals(self.database.patternSupport([2, 3, 5]), 0.2)
def test_removeTransactions(self):
self.database.removeTransactions([3, 5])
self.assertEquals(len(self.database), 3)
def test_cleanAndPrune(self):
# Not needed anymore? Add tests later
pass
from DDPMiner import DDPMine
from fptree import FPTree
from fptree.FPTree import *
from TransactionDatabase import TransactionDatabase
from optparse import OptionParser
import time
import codecs
# def run_ddpmine():
# # Just some placeholder data
# miner = DDPMine(debug=False)
# miner.mine()
if __name__ == "__main__":
database = TransactionDatabase.loadFromFile("./data/train_adt.csv", ['97'],
100)
data = TransactionDatabase.loadFromFile("./data/train_adt.csv", ['97'], 1)
data1 = TransactionDatabase.loadFromFile("./data/test_adt.csv", ['97'], 1)
# database.cleanAndPrune(2)
# print ("Cleaned database:")
# for transaction in database.transactions:
# print(str(transaction.label))
# print ("\nItems in FP tree and corresponding nodes:")
tree = FPTree()
for t in database:
tree.add(t)
# print(str(tree))
miner = DDPMine(debug=True)
start = time.clock()
Pt = miner.mine(database, 100)
