def cal_ag_cube_param(param,countNum,point_list):
laplace_num, act_num, count = main(param,countNum,point_list)
AG_CUBE_RE = cal_abs_rate(laplace_num, act_num, count)
# AG_RE = cal_abs_rate(AG_laplace_num,AG_act_num,count)
logging.warning("查询区域：x1-x2:%s-%s, y2-y2:%s-%s === AG_CUBE-相对误差为：%s" % (param.x1,param.x2,param.y1,param.y2,AG_CUBE_RE))
return AG_CUBE_RE
def load_data(filename,num):
file = open(filename)
count = 0 # 统计有多少个数据点
point_list = []
x_list = []
y_list = []
while True:
line = file.readline()
if not line:
break
elif count == num:
break
else:
str_temp = line.encode("utf-8").split()
lat = str_temp[2] # y 轴
lon = str_temp[3] # x 轴
x_list.append(lon)
y_list.append(lat)
# point = Point(float(lon),float(lat))
# point_list.append(point)
count = count + 1
logging.warning("共读取 %s 个点" % count)
# return count,point_list
return x_list,y_list
def search_square_actual_point_num(x1, x2, y1, y2, point_list):
count = 0
for item in point_list:
x_loc = item.x
y_loc = item.y
if x_loc >= x1 and x_loc <= x2 and y_loc >= y1 and y_loc <= y2:
count = count + 1
if count % 10000 == 0:
logging.warning("实际统计数量：%s" % count)
return count
def cal_ag_cube():
start_temp = time.time()
starttime = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(start_temp))
logging.warning("Start loading data, %s" % starttime)
count, point_list = load_data("simple.txt", read_num)
laplace_num, act_num, count = main(test_query_square,count, point_list)
RE = cal_abs_rate(laplace_num, act_num, count)
# AG_RE = cal_abs_rate(AG_laplace_num, AG_act_num, count)
end_temp = time.time()
endtime = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(end_temp))
logging.warning("共花费时间：%s" % (end_temp - start_temp))
def main(param,count,point_list):
# 加载经纬度集合 count:加载点的个数 point_list:点的集合
# count,point_list = load_data("simple.txt",read_num)
square_list = [] # 区域集合
# 第一次划分网格个数
m1 = cal_grid_num(count, 0.1, 10)
logging.warning("第一次划分网格个数为：%s" % m1)
# 第一次划分网格
divide_grid(square_list, int(m1), origin_square[0], origin_square[1])
logging.warning("第一次划分网格区域数量：%d " % (len(square_list)))
# 统计各个区域点的个数
cal_point_num(square_list, point_list)
# 第二次划分网格
for item in square_list:
divide_grid2(item)
# 统计第二次划分网络各个区域中的个数
for item in square_list:
cal_point_num(item.square_col,point_list)
logging.warning("=========== Starting AG ===========")
AG_count_laplace_num = {"num":0}
for item in square_list:
if len(item.square_col) == 0:
search_square_laplace_point_num(param.x1, param.x2, param.y1, param.y2, item, AG_count_laplace_num, param)
else:
for ii in item.square_col:
search_square_laplace_point_num(param.x1, param.x2, param.y1, param.y2, ii, AG_count_laplace_num, param)
AG_act_num = search_square_actual_point_num(param.x1, param.x2, param.y1, param.y2, point_list)
logging.warning("AG----actual point number is: %s" % AG_act_num)
logging.warning("AG----add laplace point number is: %s" % AG_count_laplace_num["num"])
logging.warning("========== Starting AG_CUBE =========")
# 计算 H 判断区域是否需要继续划分
for attr1 in square_list:
for attr2 in attr1.square_col:
# 划分网格
# if len(attr2.square_col) == 0:
# continue
divide_grid2(attr2)
# 统计各个网格的分割区域个数
cal_point_num(attr2.square_col, point_list)
# 生成 遍历线
genrate_line_x(attr2)
genrate_line_y(attr2)
# 判断是否需要继续划分，如果需要 返回 True, 不需要 返回 False
# 第一个参数 设定的阈值
cal_valueOfH(e, attr2)
# 如果 flag = True, 对该区域继续分割, 并找出最分割最平均的分割线
if attr2.flag:
find_gold_line(attr2,point_list)
recursion_merge(attr2.square_D1,point_list)
recursion_merge(attr2.square_D2,point_list)
else:
attr2.square_col = None
# 计算查询区域的点的个数 拉普拉斯
count_num_square = 0
count_laplace_num = {"num":0}
for item in square_list:
if len(item.square_col) == 0:
search_square_laplace_point_num(param.x1, param.x2, param.y1, param.y2, item, count_laplace_num, param)
else:
for attr1 in item.square_col:
search_square_laplace_point_num(param.x1, param.x2, param.y1, param.y2, attr1,count_laplace_num,param)
count_num_square = count_num_square + 1
if count_num_square % 10 == 0:
pass
# logging.warning("已统计区域个数为：%s" % count_num_square)
# ？
logging.warning("AG_CUBE----add laplace point number is: %s" % count_laplace_num["num"])
# 计算查询区域实际点的个数
# act_num = search_square_actual_point_num(param.x1, param.x2, param.y1, param.y2, point_list)
logging.warning("AG_CUBE----actual point number is: %s" % AG_act_num)
logging.warning("查询区域范围：x1-x2:%s-%s,y1-y2:%s-%s" % (param.x1, param.x2, param.y1, param.y2))
return count_laplace_num["num"],AG_act_num,count
def main(temp_list):
count_points, point_list = load_data("simple.txt", read_num)
start_temp = time.time()
starttime = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(start_temp))
logging.warning("Start loading data, %s" % starttime)
result = {}
AG_result = {}
get_query_square_list(temp_list)
for item in temp_list:
RE_list = []
RE_list_avg = []
count = 0
RE_AVG = 0
for i in temp_list[item]:
square_name = ("x1-x2:" + str(i.x1) + str(i.x2) + ",y1-y2:" +
str(i.y1) + str(i.y2))
temp_squ = {}
AG_temp_squ = {}
# AG_RE, RE = threading.Thread(target=cal_ag_cube_param,args=(i,count_points,point_list,))
RE = cal_ag_cube_param(i, count_points, point_list)
temp_squ[square_name] = RE
RE_AVG = RE_AVG + RE
count = count + 1
RE_list.append(temp_squ)
# print RE_AVG
AG_CUBE_AVG = RE_AVG / float(count + 1)
logging.warning("AG_CUBE====查询面积大小：%s，平均相对误差为：%s" %
(item, AG_CUBE_AVG))
logging.warning("=============================")
result[item] = AG_CUBE_AVG
# print result
# paint(result, AG_result)
result_ag = {}
for item in temp_list:
RE_list = []
RE_list_avg = []
RE_AVG = float(0.0)
count = 0
for i in temp_list[item]:
square_name = ("x1-x2:" + str(i.x1) + str(i.x2) + ",y1-y2:" +
str(i.y1) + str(i.y2))
temp_squ = {}
RE = cal_ag_param(i, count_points, point_list)
logging.warning("面积大小：%s，区域：x1-%s x2-%s y1-%s y2-%s, 相对误差为：%s" %
(item, i.x1, i.x2, i.y1, i.y2, RE))
temp_squ[square_name] = RE
RE_AVG = RE_AVG + RE
count = count + 1
RE_list.append(temp_squ)
AVG = RE_AVG / float(count)
logging.warning("面积大小：%s，平均相对误差为：%s" % (item, AVG))
result_ag[item] = AVG
logging.warning("ag_cube:%s" % result)
logging.warning("ag:%s" % result_ag)
end_temp = time.time()
endtime = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(end_temp))
logging.warning("End loading data..., %s" % endtime)
logging.warning("共花费时间：%s" % (end_temp - start_temp))
print result_ag
print result
# paint(result, AG_result)
paint(result, result_ag)