def init(data, k, QI_num=-1): #data:30162*9
"""
reset global variables
"""
global GL_K, RESULT, QI_LEN, QI_DICT, QI_RANGE, QI_ORDER
if QI_num <= 0:
QI_LEN = len(data[0]) - 1 #8,代表有幾個attribute
else:
QI_LEN = QI_num
GL_K = k
RESULT = []
# static values
QI_DICT = [] #每個domain的值的對照表
QI_ORDER = [] #每個attribute的domain
QI_RANGE = [] #每個attribute的最大值減最小值
att_values = [] #每個attribute的domain
for i in range(QI_LEN):
att_values.append(
set()
) #att_values = [set(), set(), set(), set(), set(), set(), set(), set()]
QI_DICT.append(dict()) #QI_DICT = [{}, {}, {}, {}, {}, {}, {}, {}]
for record in data:
for i in range(QI_LEN):
att_values[i].add(record[i])
for i in range(QI_LEN):
value_list = list(att_values[i])
value_list.sort(key=cmp_to_key(cmp_value))
QI_RANGE.append(value(value_list[-1]) - value(value_list[0]))
QI_ORDER.append(list(value_list))
for index, qi_value in enumerate(value_list):
QI_DICT[i][qi_value] = index
def init(data, k, QI_num=-1):
"""
reset global variables
"""
global GL_K, RESULT, QI_LEN, QI_DICT, QI_RANGE, QI_ORDER
if QI_num <= 0:
QI_LEN = len(data[0]) - 1
else:
QI_LEN = QI_num
GL_K = k
RESULT = []
# static values
QI_DICT = []
QI_ORDER = []
QI_RANGE = []
att_values = []
for i in range(QI_LEN):
att_values.append(set())
QI_DICT.append(dict())
for record in data:
for i in range(QI_LEN):
att_values[i].add(record[i])
for i in range(QI_LEN):
value_list = list(att_values[i])
value_list.sort(key=cmp_to_key(cmp_value))
QI_RANGE.append(value(value_list[-1]) - value(value_list[0]))
QI_ORDER.append(list(value_list))
for index, qi_value in enumerate(value_list):
QI_DICT[i][qi_value] = index
def init(data, k, QI_num=-1):
"""
reset global variables
"""
global GL_K, RESULT, QI_LEN, QI_DICT, QI_RANGE, QI_ORDER
if QI_num <= 0:
QI_LEN = len(data[0]) - 1
else:
QI_LEN = QI_num
GL_K = k
RESULT = []
# static values
QI_DICT = []
QI_ORDER = []
QI_RANGE = []
att_values = []
for i in range(QI_LEN):
att_values.append(set())
QI_DICT.append(dict())
for record in data:
for i in range(QI_LEN):
att_values[i].add(record[i])
for i in range(QI_LEN):
value_list = list(att_values[i])
value_list.sort(key=cmp_to_key(cmp_value))
QI_RANGE.append(value(value_list[-1]) - value(value_list[0]))
QI_ORDER.append(list(value_list))
for index, qi_value in enumerate(value_list):
QI_DICT[i][qi_value] = index
def get_normalized_width(partition, index):
"""
return Normalized width of partition
similar to NCP
"""
d_order = QI_ORDER[index]
width = value(d_order[partition.high[index]]) - value(d_order[partition.low[index]])
if width == QI_RANGE[index]:
return 1
return width * 1.0 / QI_RANGE[index]
def get_normalized_width(partition, index):
"""
return Normalized width of partition
similar to NCP
"""
d_order = QI_ORDER[index]
width = value(d_order[partition.high[index]]) - value(
d_order[partition.low[index]])
if width == QI_RANGE[index]:
return 1
return width * 1.0 / QI_RANGE[index]