def wsize_avgsk():
print("=== Test how window size affect candidate skyline ===")
wsize = [100, 200, 300, 400, 500, 600, 700, 800, 900, 1000]
dqueue = batchImport('10000_dim2_pos5_rad5_01000.csv', 5)
for w in wsize:
print('========== window size = ' + str(w) + ' ==========')
print('---------- Brute force ----------')
tbsky = slideBPSky(2, 5, 5, [0, 1000], wsize=w)
avgsk1, avgsk2 = 0, 0
for i in range(10000):
tbsky.receiveData(dqueue[i])
tbsky.updateSkyline()
avgsk1 += len(tbsky.getSkyline())
avgsk2 += len(tbsky.getSkyline2())
tbsky.removeRtree()
avgsk1, avgsk2 = avgsk1 / 10000, avgsk2 / 10000
print('Avg. sky1: ' + str(avgsk1))
print('Avg. sky2: ' + str(avgsk2))
print('---------- Update ----------')
tusky = slideUPSky(2, 5, 5, [0, 1000], wsize=w)
avgsk1, avgsk2 = 0, 0
for i in range(10000):
tusky.receiveData(dqueue[i])
tusky.updateSkyline()
avgsk1 += len(tusky.getSkyline())
avgsk2 += len(tusky.getSkyline2())
tusky.removeRtree()
avgsk1, avgsk2 = avgsk1 / 10000, avgsk2 / 10000
print('Avg. sky1: ' + str(avgsk1))
print('Avg. sky2: ' + str(avgsk2))
def instance_avgsk():
print('=== Test how instance count affect candidate skyline ===')
inst = [3, 4, 5, 6, 7, 8, 9, 10]
for ins in inst:
dqueue = batchImport('10000_dim2_pos' + str(ins) + '_rad5_01000.csv',
ins)
print('========== instance count = ' + str(ins) + ' ==========')
print('---------- Brute force ----------')
tbsky = slideBPSky(2, ins, 5, [0, 1000], wsize=300)
avgsk1, avgsk2 = 0, 0
for i in range(10000):
tbsky.receiveData(dqueue[i])
tbsky.updateSkyline()
avgsk1 += len(tbsky.getSkyline())
avgsk2 += len(tbsky.getSkyline2())
tbsky.removeRtree()
avgsk1, avgsk2 = avgsk1 / 10000, avgsk2 / 10000
print('Avg. sky1: ' + str(avgsk1))
print('Avg. sky2: ' + str(avgsk2))
print('---------- Update ----------')
tusky = slideUPSky(2, ins, 5, [0, 1000], wsize=300)
avgsk1, avgsk2 = 0, 0
for i in range(10000):
tusky.receiveData(dqueue[i])
tusky.updateSkyline()
avgsk1 += len(tusky.getSkyline())
avgsk2 += len(tusky.getSkyline2())
tusky.removeRtree()
avgsk1, avgsk2 = avgsk1 / 10000, avgsk2 / 10000
print('Avg. sky1: ' + str(avgsk1))
print('Avg. sky2: ' + str(avgsk2))
def radius_avgsk():
print('=== Test how data radius affect candidate skyline ===')
radius = [3, 4, 5, 6, 7, 8, 9, 10]
for r in radius:
dqueue = batchImport('10000_dim2_pos5_rad' + str(r) + '_01000.csv', 5)
print('========== radius = ' + str(r) + ' ==========')
print('---------- Brute force ----------')
tbsky = slideBPSky(2, 5, r, [0, 1000], wsize=300)
avgsk1, avgsk2 = 0, 0
for i in range(10000):
tbsky.receiveData(dqueue[i])
tbsky.updateSkyline()
avgsk1 += len(tbsky.getSkyline())
avgsk2 += len(tbsky.getSkyline2())
tbsky.removeRtree()
avgsk1, avgsk2 = avgsk1 / 10000, avgsk2 / 10000
print('Avg. sky1: ' + str(avgsk1))
print('Avg. sky2: ' + str(avgsk2))
print('---------- Update ----------')
tusky = slideUPSky(2, 5, r, [0, 1000], wsize=300)
avgsk1, avgsk2 = 0, 0
for i in range(10000):
tusky.receiveData(dqueue[i])
tusky.updateSkyline()
avgsk1 += len(tusky.getSkyline())
avgsk2 += len(tusky.getSkyline2())
tusky.removeRtree()
avgsk1, avgsk2 = avgsk1 / 10000, avgsk2 / 10000
print('Avg. sky1: ' + str(avgsk1))
print('Avg. sky2: ' + str(avgsk2))
def dim_avgsk():
print("=== Test how dimension of data affect candidate size ===")
dim = [2, 3, 4, 5, 6, 7, 8, 9, 10]
for d in dim:
dqueue = batchImport('10000_dim'+str(d)+'_pos5_rad5_01000.csv', 5)
print('========== Data dimension = '+ str(d) + ' ==========')
print('---------- Brute force ----------')
tbsky = slideBPSky(d, 5, 5, [0,1000], wsize=300)
avgsk1, avgsk2 = 0, 0
for i in range(10000):
tbsky.receiveData(dqueue[i])
tbsky.updateSkyline()
avgsk1 += len(tbsky.getSkyline())
avgsk2 += len(tbsky.getSkyline2())
tbsky.removeRtree()
avgsk1, avgsk2 = avgsk1/10000, avgsk2/10000
print('Avg. sky1: '+ str(avgsk1))
print('Avg. sky2: '+ str(avgsk2))
print('---------- Update ----------')
tusky = slideUPSky(d, 5, 5, [0,1000], wsize=300)
avgsk1, avgsk2 = 0, 0
for i in range(10000):
tusky.receiveData(dqueue[i])
tusky.updateSkyline()
avgsk1 += len(tusky.getSkyline())
avgsk2 += len(tusky.getSkyline2())
tusky.removeRtree()
avgsk1, avgsk2 = avgsk1/10000, avgsk2/10000
print('Avg. sky1: '+ str(avgsk1))
print('Avg. sky2: '+ str(avgsk2))
def radius_time():
print('=== Test how data radius affect running time ===')
radius = [3, 4, 5, 6, 7, 8, 9, 10]
for r in radius:
dqueue = batchImport('10000_dim2_pos5_rad' + str(r) + '_01000.csv', 5)
print('========== radius = ' + str(r) + ' ==========\n')
print('---------- Brute force ----------')
tbsky = slideBPSky(2, 5, r, [0, 1000], wsize=300)
start_time = time.time()
for i in range(10000):
tbsky.receiveData(dqueue[i])
tbsky.updateSkyline()
tbsky.removeRtree()
print("--- %s seconds ---" % (time.time() - start_time))
print('---------- Update ----------')
tusky = slideUPSky(2, 5, r, [0, 1000], wsize=300)
start_time = time.time()
for i in range(10000):
tusky.receiveData(dqueue[i])
tusky.updateSkyline()
tusky.removeRtree()
print("--- %s seconds ---" % (time.time() - start_time))
def wsize_time():
print("=== Test how window size affect running time ===")
wsize = [100, 200, 300, 400, 500, 600, 700, 800, 900, 1000]
dqueue = batchImport('10000_dim2_pos5_rad5_01000.csv', 5)
for w in wsize:
print('========== window size = ' + str(w) + ' ==========')
print('---------- Brute force ----------')
tbsky = slideBPSky(2, 5, 5, [0, 1000], wsize=w)
start_time = time.time()
for i in range(10000):
tbsky.receiveData(dqueue[i])
tbsky.updateSkyline()
tbsky.removeRtree()
print("--- %s seconds ---" % (time.time() - start_time))
print('---------- Update ----------')
tusky = slideUPSky(2, 5, 5, [0, 1000], wsize=w)
start_time = time.time()
for i in range(10000):
tusky.receiveData(dqueue[i])
tusky.updateSkyline()
tusky.removeRtree()
print("--- %s seconds ---" % (time.time() - start_time))
def dim_time():
print("=== Test how dimension of data affect running time ===")
dim = [2, 3, 4, 5, 6, 7, 8, 9, 10]
for d in dim:
dqueue = batchImport('10000_dim'+str(d)+'_pos5_rad5_01000.csv', 5)
print('========== Data dimension = '+ str(d) + ' ==========')
print('---------- Brute force ----------')
tbsky = slideBPSky(d, 5, 5, [0,1000], wsize=300)
start_time = time.time()
for i in range(10000):
tbsky.receiveData(dqueue[i])
tbsky.updateSkyline()
tbsky.removeRtree()
print("--- %s seconds ---" % (time.time() - start_time))
print('---------- Update ----------')
tusky = slideUPSky(d, 5, 5, [0,1000], wsize=300)
start_time = time.time()
for i in range(10000):
tusky.receiveData(dqueue[i])
tusky.updateSkyline()
tusky.removeRtree()
print("--- %s seconds ---" % (time.time() - start_time))
def instance_time():
print('=== Test how instance count affect running time ===')
inst = [3, 4, 5, 6, 7, 8, 9, 10]
for ins in inst:
dqueue = batchImport('10000_dim2_pos' + str(ins) + '_rad5_01000.csv',
ins)
print('========== instance count = ' + str(ins) + ' ==========')
print('---------- Brute force ----------')
tbsky = slideBPSky(2, ins, 5, [0, 1000], wsize=300)
start_time = time.time()
for i in range(10000):
tbsky.receiveData(dqueue[i])
tbsky.updateSkyline()
tbsky.removeRtree()
print("--- %s seconds ---" % (time.time() - start_time))
print('---------- Update ----------')
tusky = slideUPSky(2, ins, 5, [0, 1000], wsize=300)
start_time = time.time()
for i in range(10000):
tusky.receiveData(dqueue[i])
tusky.updateSkyline()
tusky.removeRtree()
print("--- %s seconds ---" % (time.time() - start_time))
# Connect to server and send data
sock.connect((self.host, self.port))
sock.sendall(sdata)
# received = str(sock.recv(1024), "utf-8")
print("Sent: {}".format(updateInfo))
# print("Received: {}".format(received))
if __name__ == "__main__":
config = configparser.ConfigParser()
config.read('edge.config')
PORT = int(config['DEFAULT'].get('Port'))
HOST = config['DEFAULT'].get('Host')
edge = Edge(HOST, PORT)
usky = slideUPSky(2, 5, 4, [0, 1000], wsize=10)
dqueue = batchImport('1500_dim2_pos4_rad5_01000.csv', 4)
for i in range(15):
oldsk = usky.getSkyline().copy()
oldsk2 = usky.getSkyline2().copy()
usky.receiveData(dqueue[i])
out = usky.getOutdated().copy()
usky.updateSkyline()
usk1 = list(set(usky.getSkyline()) - set(oldsk))
usk2 = list(set(usky.getSkyline2()) - set(oldsk2))
result = {'Delete': out, 'SK1': usk1, 'SK2': usk2}
# edge.sendSK(usky.getSkyline(),usky.getSkyline2())
edge.send(result)
print('SK1')
for d in usky.getSkyline():
import os, sys
sys.path.append(os.path.abspath(os.pardir))
import pickle
import time
from data.dataClass import Data, batchImport
from skyline.slideUPSky import slideUPSky
from visualize import visualize
if __name__ == "__main__":
eid = input("Edge id: ")
usky = slideUPSky(2, 5, 5, [0, 1000], wsize=300)
dqueue = batchImport('10000_dim2_pos5_rad5_01000.csv', 5)
idx = [i for i in range(10000) if i % 2 == 1]
with open('pickle_edge' + eid + '.pickle', 'wb') as f:
start_time = time.time()
for i in idx:
oldsk = usky.getSkyline().copy()
oldsk2 = usky.getSkyline2().copy()
usky.receiveData(dqueue[i])
out = usky.getOutdated().copy()
usky.updateSkyline()
usk1 = list(set(usky.getSkyline()) - set(oldsk))
usk2 = list(set(usky.getSkyline2()) - set(oldsk2))
result = {'Delete': out, 'SK1': usk1, 'SK2': usk2}
pickle.dump(result, f)
print("--- %s seconds ---" % (time.time() - start_time))
# Sliding window update PSky
from skyline.PSky import PSky
from data.dataClass import Data, batchImport
from visualize.visualize import visualize
import random
from matplotlib import pyplot as plt
from matplotlib.patches import Rectangle
from data.dataClass import Data, batchImport
from skyline.slideUPSky import slideUPSky
test = slideUPSky(2, 5, 4, [0,100], wsize=10)
data = batchImport('20_dim2_pos5_rad5_0100.csv', 5)
fig = plt.figure()
ax = fig.add_subplot(111)
for d in data:
x = [d.getLocation(i)[0] for i in range(5)]
y = [d.getLocation(j)[1] for j in range(5)]
ax.scatter(x,y,alpha=0.5)
ax.add_patch(Rectangle(xy=(d.getLocationMin()[0],d.getLocationMin()[1]),width=d.getLocationMax()[0]-d.getLocationMin()[0], height=d.getLocationMax()[1]-d.getLocationMin()[1], linewidth=1, fill=False))
plt.show()
# for i in range(100):
# test.receiveData(dqueue[i])
# test.updateSkyline()