def test_RMT_with_merge_case(self):
init_tree()
att_trees = [ATT_TREE, ATT_TREE]
data = [['a2', ['a1', 'b1']],
['a1', ['a1', 'b1']],
['b2', ['a2', 'b2']],
['b2', ['a2', 'b2']],
['b1', ['a1', 'b1']],
['b1', ['a1', 'b1']]]
_, result = rt_anon(att_trees, data, 'RMT', 2, 2, 0.5)
self.assertTrue(abs(result[0] - 50) <= 0.001)
self.assertTrue(abs(result[1] - 0) <= 0.001)
_, result = rt_anon(att_trees, data, 'RMT', 2, 2, 0.7)
self.assertTrue(abs(result[0] - 200.0 / 3) <= 0.001)
self.assertTrue(abs(result[1] - 0) <= 0.001)
def get_result_m(att_tree, data, type_alg, k=DEFAULT_K, threshold=DEFAULT_T):
"""
change k, whle fixing size of dataset
"""
print "K=%d" % k
print "Threshold=%.2f" % threshold
print "Size of Data", len(data)
data_back = copy.deepcopy(data)
# for m in range(1, 100, 5):
all_rncp = []
all_tncp = []
all_rtime = []
m_range = [1, 2, 3, 4, 5, M_MAX]
for m in m_range:
print '#' * 30
print "m=%d" % m
result, eval_result = rt_anon(att_tree, data, type_alg, k, m, threshold)
# save_to_file((att_tree, data, result, k, m))
data = copy.deepcopy(data_back)
print "RNCP %0.2f" % eval_result[0] + "%"
all_rncp.append(round(eval_result[0], 2))
print "TNCP %0.2f" % eval_result[1] + "%"
all_tncp.append(round(eval_result[1], 2))
print "Running time %0.2f" % eval_result[2] + " seconds"
all_rtime.append(round(eval_result[2], 2))
print "m range", m_range
print "RNCP", all_rncp
print "TNCP", all_tncp
print "Running time", all_rtime
def get_result_k(att_tree, data, type_alg, m=DEFAULT_M, threshold=DEFAULT_T):
"""
change k, whle fixing size of dataset
"""
data_back = copy.deepcopy(data)
# for k in range(5, 105, 5):
print "m=%d" % m
print "Threshold=%.2f" % threshold
print "Size of Data", len(data)
all_rncp = []
all_tncp = []
all_rtime = []
# for k in range(5, 55, 5):
# if k in [2, 5, 10, 25, 50, 100]:
# continue
k_range = [2, 5, 10, 25, 50, 100]
for k in k_range:
print '#' * 30
print "K=%d" % k
result, eval_result = rt_anon(att_tree, data, type_alg, k, m, threshold)
# save_to_file((att_tree, data, result, k, m))
data = copy.deepcopy(data_back)
print "RNCP %0.2f" % eval_result[0] + "%"
all_rncp.append(round(eval_result[0], 2))
print "TNCP %0.2f" % eval_result[1] + "%"
all_tncp.append(round(eval_result[1], 2))
print "Running time %0.2f" % eval_result[2] + " seconds"
all_rtime.append(round(eval_result[2], 2))
print "K range", k_range
print "RNCP", all_rncp
print "TNCP", all_tncp
print "Running time", all_rtime
def test_RMR_RMT_with_self_case(self):
init_tree()
att_trees = [ATT_TREE, ATT_TREE]
data = [['a1', ['a1', 'b1', 'b2']],
['a1', ['a2', 'b1']],
['a2', ['a2', 'b1', 'b2']],
['a2', ['a1', 'a2', 'b2']]]
_, result = rt_anon(att_trees, data, 'RMR', 2, 2, 1.0)
self.assertTrue(abs(result[1] - 0.5 * 5 / 11 * 100) <= 0.001)
_, result = rt_anon(att_trees, data, 'RMR', 2, 2, 0.1)
self.assertTrue(abs(result[0] - 0) <= 0.001)
self.assertTrue(abs(result[1] - 50) <= 0.001)
_, result = rt_anon(att_trees, data, 'RMT', 2, 2, 1.0)
self.assertTrue(abs(result[1] - 0.5 * 5 / 11 * 100) <= 0.001)
_, result = rt_anon(att_trees, data, 'RMT', 2, 2, 0.1)
self.assertTrue(abs(result[0] - 0) <= 0.001)
self.assertTrue(abs(result[1] - 50) <= 0.001)
def test_RMR_with_full_threshold(self):
init_tree()
att_trees = [ATT_TREE, ATT_TREE]
data = [['a1', ['a1']],
['a1', ['a1', 'a2']],
['b1', ['b1', 'b2']],
['b2', ['b1', 'b2']],
['b1', ['a1', 'a2', 'b2']],
['b2', ['a1', 'a2', 'b2']],
['a1', ['a1', 'a2', 'b1', 'b2']]]
_, result = rt_anon(att_trees, data, 'RMR', 2, 2, 1.0)
self.assertTrue(abs(result[1] - 8 * 0.5 * 100 / 17) <= 0.001)
def get_result_one(att_tree, data, type_alg, k=DEFAULT_K, m=DEFAULT_M, threshold=DEFAULT_T):
"""
run RT_ANON for one time, with k=10
"""
print "K=%d" % k
print "Size of Data", len(data)
print "m=%d" % m
print "Threshold=%.2f" % threshold
result, eval_result = rt_anon(att_tree, data, type_alg, k, m, threshold)
# save_to_file((att_tree, data, result, k, m))
print "RNCP %0.2f" % eval_result[0] + "%"
print "TNCP %0.2f" % eval_result[1] + "%"
print "Running time %0.2f" % eval_result[2] + " seconds"
def get_result_dataset(att_tree, data, type_alg='RMR',
k=DEFAULT_K, m=DEFAULT_M, threshold=DEFAULT_T, num_test=10):
"""
fix k, while changing size of dataset
num_test is the test nubmber.
"""
print "K=%d" % k
print "m=%d" % m
print "Threshold=%.2f" % threshold
data_back = copy.deepcopy(data)
length = len(data_back)
joint = 5000
datasets = []
check_time = length / joint
if length % joint == 0:
check_time -= 1
for i in range(check_time):
datasets.append(joint * (i + 1))
# datasets.append(length)
all_rncp = []
all_tncp = []
all_rtime = []
for pos in datasets:
rncp = tncp = rtime = 0
if pos > length:
continue
print '#' * 30
print "size of dataset %d" % pos
for j in range(num_test):
temp = random.sample(data, pos)
result, eval_result = rt_anon(att_tree, temp, type_alg, k, m, threshold)
# save_to_file((att_tree, temp, result, k, m), number=j)
rncp += eval_result[0]
tncp += eval_result[1]
rtime += eval_result[2]
data = copy.deepcopy(data_back)
rncp /= num_test
tncp /= num_test
rtime /= num_test
print "RNCP %0.2f" % rncp + "%"
all_rncp.append(round(rncp, 2))
print "TNCP %0.2f" % tncp + "%"
all_tncp.append(round(tncp, 2))
print "Running time %0.2f" % rtime + " seconds"
all_rtime.append(round(rtime, 2))
print "Size of datasets", datasets
print "RNCP", all_rncp
print "TNCP", all_tncp
print "Running time", all_rtime
def test_RMR_with_merge_case(self):
init_tree()
att_trees = [ATT_TREE, ATT_TREE]
data = [['a1', ['a1', 'b1', 'b2']],
['a1', ['a2', 'b1']],
['a2', ['a2', 'b1', 'b2']],
['a2', ['a1', 'a2', 'b2']],
['b1', ['a2', 'b1', 'b2']],
['b1', ['a1', 'a2', 'b2']],
['b2', ['a2', 'b1', 'b2']],
['b2', ['a1', 'a2', 'b2']]]
_, result = rt_anon(att_trees, data, 'RMR', 2, 2, 0.5)
# print result
self.assertTrue(abs(result[0] - 50) <= 0.001)
self.assertTrue(abs(result[1] - 0.5 * 5 / 23 * 100) <= 0.001)
