def get_result_missing(att_trees, data, k=DEFAULT_K, n=10):
"""
change nubmber of missing, whle fixing k, qi and size of dataset
"""
data_back = copy.deepcopy(data)
length = len(data_back)
qi_len = len(data[0]) - 1
raw_missing = raw_missing_record = 0
print "K=%d" % k
for record in data:
flag = False
for value in record:
if value == '*':
raw_missing += 1
flag = True
if flag:
raw_missing_record += 1
# print "Missing Percentage %.2f" % (raw_missing * 100.0 / (length * qi_len)) + '%%'
# each evaluation varies add 5% missing values
check_percentage = [5, 10, 25, 50, 75]
datasets = []
for p in check_percentage:
joint = int(0.01 * p * length * qi_len) - raw_missing
datasets.append(joint)
all_ncp = []
all_rtime = []
all_pollution = []
for i, joint in enumerate(datasets):
ncp = rtime = pollution = 0.0
for j in range(n):
gen_missing_dataset(data, joint)
if __DEBUG:
missing_rate(data)
_, eval_result = semi_partition(att_trees, data, k)
data = copy.deepcopy(data_back)
ncp += eval_result[0]
rtime += eval_result[1]
pollution += eval_result[2]
ncp /= n
rtime /= n
pollution /= n
if __DEBUG:
print "check_percentage", check_percentage[i]
print "Add missing %d" % joint
print "Average NCP %0.2f" % ncp + "%"
print "Running time %0.2f" % rtime + "seconds"
print "Missing Pollution = %.2f" % pollution + "%"
print '#' * 30
all_ncp.append(round(ncp, 2))
all_rtime.append(round(rtime, 2))
all_pollution.append(round(pollution, 2))
print "All NCP", all_ncp
print "All Running time", all_rtime
print "Missing Pollution", all_pollution
print '#' * 30
def get_result_one(att_trees, data, k=DEFAULT_K):
"run mondrian for one time, with k=10"
print "K=%d" % k
data_back = copy.deepcopy(data)
missing_rate(data)
_, eval_result = mondrian_delete_missing(att_trees, data, k)
print "Mondrian"
print "NCP %0.2f" % eval_result[0] + "%"
print "Running time %0.2f" % eval_result[1] + "seconds"
print "Missing Pollution = %.2f %%" % eval_result[2]
data = copy.deepcopy(data_back)
_, eval_result = mondrian(att_trees, data, k)
print "Enhanced Mondrian"
print "NCP %0.2f" % eval_result[0] + "%"
print "Running time %0.2f" % eval_result[1] + "seconds"
print "Missing Pollution = %.2f %%" % eval_result[2]
def get_result_one(att_trees, data, k=DEFAULT_K):
"run mondrian for one time, with k=10"
print "K=%d" % k
data_back = copy.deepcopy(data)
missing_rate(data)
_, eval_result = mondrian_delete_missing(att_trees, data, k)
print "Mondrian"
print "NCP %0.2f" % eval_result[0] + "%"
print "Running time %0.2f" % eval_result[1] + "seconds"
print "Missing Pollution = %.2f %%" % eval_result[2]
data = copy.deepcopy(data_back)
_, eval_result = mondrian(att_trees, data, k)
print "Enhanced Mondrian"
print "NCP %0.2f" % eval_result[0] + "%"
print "Running time %0.2f" % eval_result[1] + "seconds"
print "Missing Pollution = %.2f %%" % eval_result[2]
def get_result_missing(att_trees, data, k=DEFAULT_K, n=DEFAULT_K):
"""
change nubmber of missing, whle fixing k, qi and size of dataset
"""
data_back = copy.deepcopy(data)
length = len(data_back)
qi_len = len(data[0]) - 1
raw_missing = raw_missing_record = 0
print "K=%d" % k
for record in data:
flag = False
for value in record:
if value == '*':
raw_missing += 1
flag = True
if flag:
raw_missing_record += 1
# print "Missing Percentage %.2f" % (raw_missing * 100.0 / (length * qi_len)) + '%%'
# each evaluation varies add 5% missing values
check_percentage = [5, 10, 25, 50, 75]
datasets = []
for p in check_percentage:
joint = int(0.01 * p * length * qi_len) - raw_missing
datasets.append(joint)
all_ncp = []
all_rtime = []
all_pollution = []
deletion_all_ncp = []
deletion_all_rtime = []
for i, joint in enumerate(datasets):
ncp = rtime = pollution = 0.0
for j in range(n):
gen_missing_dataset(data, joint)
if __DEBUG:
missing_rate(data)
_, eval_result = mondrian(att_trees, data, k)
data = copy.deepcopy(data_back)
ncp += eval_result[0]
rtime += eval_result[1]
pollution += eval_result[2]
ncp /= n
rtime /= n
pollution /= n
if __DEBUG:
print "check_percentage", check_percentage[i]
print "Add missing %d" % joint
print "Average NCP %0.2f" % ncp + "%"
print "Running time %0.2f" % rtime + "seconds"
print "Missing Pollution = %.2f" % pollution + "%"
print '#' * 30
all_ncp.append(round(ncp, 2))
all_rtime.append(round(rtime, 2))
all_pollution.append(round(pollution, 2))
ncp = rtime = pollution = 0.0
for j in range(n):
gen_missing_dataset(data, joint)
if __DEBUG:
missing_rate(data)
_, eval_result = mondrian_delete_missing(att_trees, data, k)
data = copy.deepcopy(data_back)
ncp += eval_result[0]
rtime += eval_result[1]
ncp /= n
rtime /= n
if __DEBUG:
print "Add missing %d" % joint
print "Average NCP %0.2f" % ncp + "%"
print "Running time %0.2f" % rtime + "seconds"
print "Missing Pollution = %.2f" % pollution + "%"
print '#' * 30
deletion_all_ncp.append(round(ncp, 2))
deletion_all_rtime.append(round(rtime, 2))
print "Mondrian"
print "All NCP", deletion_all_ncp
print "All Running time", deletion_all_rtime
print "Enhanced Mondrian"
print "All NCP", all_ncp
print "All Running time", all_rtime
print "Missing Pollution", all_pollution
print '#' * 30