def test_issubsequence(self):
sub = []
sup = []
self.assertTrue(Utils.isSubsequence(sub, sup))
sub = []
sup = [[1], [2]]
self.assertTrue(Utils.isSubsequence(sub, sup))
sub = [[2], [3]]
sup = [[1, 2, 3], [1, 2, 3, 4]]
self.assertTrue(Utils.isSubsequence(sub, sup))
sub = [[1, 3], [3, 3, 3]]
sup = [[1, 2, 3], [1, 2, 3, 4]]
self.assertFalse(Utils.isSubsequence(sub, sup))
def canPrune(self, seq):
sLowestMIS = self.getStrictlyMinimumMIS(seq)
k = Utils.seqLength(seq)
for i in range(k):
item = Utils.getItem(seq, i)
if self.MS[item] == sLowestMIS:
continue
k_1_subseq = Utils.removeItem(seq, i)
count = 0
for d in self.T:
if Utils.isSubsequence(k_1_subseq, d):
count += 1
if float(count) / len(self.T) < self.getMinMIS(k_1_subseq):
return True
return False
def run(self):
L = Utils.getUniqueItems(self.T)
SUP = Utils.genItemSupportCount(L, self.T)
lSUP = {}
for l in range(len(L)):
lSUP[L[l]] = SUP[l]
if len(L) > 3:
print "SORRY! Can't run Brute Force with these large data"
return []
C = Utils.generateAllSubsets(L)
S = Utils.generateAllSequences(C)
outputData = []
for seq in S:
count = 0
minSUP = 999
maxSUP = 0
minMIS = 999
for s in range(len(seq)):
for i in seq[s]:
if lSUP[i] < minSUP:
minSUP = lSUP[i]
if lSUP[i] > maxSUP:
maxSUP = lSUP[i]
if self.MS[i] < minMIS:
minMIS = self.MS[i]
for d in self.T:
if Utils.isSubsequence(seq, d):
count += 1
if ((float(count) / len(self.T)) >= minMIS) and (float(maxSUP - minSUP) / len(self.T) <= self.SDC):
outputData.append(seq)
return outputData
data_file = sys.argv[2]
para_file = sys.argv[3]
result_file = sys.argv[4]
DP = DataProcessor(data_file, para_file, result_file, False)
inputData = DP.loadInput()
startTime = datetime.now()
print 'Execution started at:', startTime
algo = pyMSGSP(inputData["T"], inputData["MS"], inputData["SDC"], logging.INFO)
# algo = BruteForceSPM(inputData["T"], inputData["MS"], inputData["SDC"])
outputData = algo.run()
print 'Execution time:', datetime.now() - startTime
outputDict = defaultdict(list)
for seq in outputData:
count = 0
for d in inputData["T"]:
if Utils.isSubsequence(seq, d):
count += 1
outputDict[Utils.seqLength(seq)].append((seq, count))
DP.printOutput(outputDict)
