Commit 35f23453 authored by Quentin Verout's avatar Quentin Verout
Browse files

Add resulats.pdf and runTest.py

parent 53798901

distance.py

+30 −30
Original line number Diff line number Diff line
@@ -9,10 +9,9 @@ import copy 
import statistics

import collections

import operator

import textdistance

from itertools import tee







class Robustness:

	def __init__(self, pwdContainer=None):

		"""
@@ -21,6 +20,7 @@ class Robustness: 
		

		self.l = list()

		self.distance = leven.distance

		#self.distance = textdistance.jaro_winkler

		self.algo = self.random_permutation

		

		if type(pwdContainer) == str:
@@ -78,6 +78,7 @@ class Robustness: 
			k -= 1

		after = time.time()

		cost = int(c * ratio)

		cost = c

		#print("Run an excursion and its cost rate in ", after-before, "seconds")

		return cost
@@ -226,7 +227,7 @@ class Robustness: 
			word =''.join(avg_word)

		return word

		

	def k_means(self, k=2, max_iters=100, algo=2):

	def k_means(self, k=10, max_iters=100, algo=2):

		"""

		Performs k-means clustering with Levenshtein distance. This function return the cluster assigment.

		This function indicate also the number of iterations to find the optimal solutions.
@@ -244,9 +245,9 @@ class Robustness: 
		"""

		

		#Centroids

		print(self.l)

		#print(self.l)

		centroids = random.sample(self.l, k)

		print("First centroids :", centroids,"\n")

		#print("First centroids :", centroids,"\n")

		

		#Current Average

		ex_centroids = [[] for i in range(k)]
@@ -266,8 +267,8 @@ class Robustness: 
			

			#For every word	

			for w in copy_perm:

				clostest_word =''

				mini_dist = 30000

				clostest_word = ""

				mini_dist = 999999

				

				#Wich centroids is closest

				for i in centroids:
@@ -280,33 +281,29 @@ class Robustness: 
			ex_centroids = copy.deepcopy(centroids)

			

			#Update centroids (middle of each cluster)

			print("Cluster", cluster)

			#print("Cluster", cluster)

			for i in range (len(centroids)):

				centroids[i] = self.avg_wordlist(cluster[i])

			#print("Update centroids :", centroids,"\n")

					

			print("Update centroids :", centroids,"\n")

			#Converge ? If yes, then break

			if ex_centroids == centroids or y == max_iters-1:

				print("Breaked at the %dth iterations" % (y))

				

				cost = 0

			

				for i in cluster:

					if algo == 1:

						cost += Robustness(i).cost_rate()

						print("cout", cost)

					else :

						closest = Robustness(i).closest_word()

						second_cost = 0

						for i in closest:

							second_cost = i[1]

						cost += second_cost

					

			#Converge ? If yes, then break

			if ex_centroids == centroids:

				print("Breaked at the %dth iterations" % (y))

				return cluster, cost

					

		

		print("Optimal solutions not found after %d th iterations" %(y))

		print("cost", cost)

				return cluster, cost

			print(y)

    

	def sampling(self, t, algo=1):

		"""
@@ -367,8 +364,11 @@ if __name__ == '__main__': 
	

	before = time.time()

	

	cluster = R.k_means()

	print("yo",cluster)

	l=[]

	for i in range(10):

		l.append(R.min_cost())

	l = sorted(l)

	print(l[int(len(l)/2)])

	

	after = time.time()

	print("One complete random_permutation in %d seconds" % (after-before))

resultats.pdf

0 → 100644
+85.7 KiB

File added.

No diff preview for this file type.

View file

runTest.py

0 → 100644
+216 −0
Original line number Diff line number Diff line 
#!/usr/bin/env python3



import sys

import time

from distance import *



if __name__ == '__main__':

	

	print("\nThis test will run some function on two database to compare their robustness, The robustness of a database depends on its length, this is why the minimum length of the two files will be the maximum reference for both files\n")

	before = time.time()

	firstFile = Robustness(sys.argv[1])

	

		

	after = time.time()

	print("First file loaded in %d seconds." % (after-before))

	

	before = time.time()

	secondFile = Robustness(sys.argv[2])

	after = time.time()

	print("Second file loaded in %d seconds.\n" % (after-before),)

	

	mini_D = min(len(firstFile.l), len(secondFile.l))

	firstFile.k, secondFile.k = mini_D, mini_D

	print("Maximum D =", mini_D)

	

	res = []

	

	print("\n=============================================================\n")

	

	print("Function : COST_RATE")

	

	firstFile.k, secondFile.k = 1000, 1000

	

	l1, l2 = [], []

	

	for i in range(100):

		l1.append(firstFile.cost_rate())

		l2.append(secondFile.cost_rate())

		l1 = sorted(l1)

		l2 = sorted(l2)

	

	print("d =", firstFile.k," first file got a median of :", l1[int(len(l1)/2)], "while the second one got a median of: " , l2[int(len(l2)/2)], "		(100 iterations)")

	res.append(l1[int(len(l1)/2)]>l2[int(len(l2)/2)])

	

	firstFile.k, secondFile.k = 1000000, 1000000

	

	l1, l2 = [], []

	

	for i in range(100):

		l1.append(firstFile.cost_rate())

		l2.append(secondFile.cost_rate())

		l1 = sorted(l1)

		l2 = sorted(l2)

	

	print("d =", firstFile.k," first file got a median of :", l1[int(len(l1)/2)], "while the second one got a median of: " , l2[int(len(l2)/2)], "		(100 iterations)")

	res.append(l1[int(len(l1)/2)]>l2[int(len(l2)/2)])

	

	firstFile.k, secondFile.k = mini_D, mini_D

	

	l1, l2 = [], []

	

	for i in range(10):

		l1.append(firstFile.cost_rate())

		l2.append(secondFile.cost_rate())

		l1 = sorted(l1)

		l2 = sorted(l2)

	

	print("d =", firstFile.k," first file got a median of :", l1[int(len(l1)/2)], "while the second one got a median of: " , l2[int(len(l2)/2)], "		(10 iterations)")

	res.append(l1[int(len(l1)/2)]>l2[int(len(l2)/2)])

	

	print("\n=============================================================\n")

	

	print("Function : MIN_COST")



	firstFile.k, secondFile.k = 1000, 1000

	

	l1, l2 = [], []

	

	for i in range(100):

		l1.append(firstFile.min_cost())

		l2.append(secondFile.min_cost())

		l1 = sorted(l1)

		l2 = sorted(l2)

	

	print("d =", firstFile.k," first file got a median of :", l1[int(len(l1)/2)], "while the second one got a median of: " , l2[int(len(l2)/2)], "		(100 iterations)")

	res.append(l1[int(len(l1)/2)]>l2[int(len(l2)/2)])

	

	firstFile.k, secondFile.k = 100000, 100000

	

	l1, l2 = [], []

	

	for i in range(10):

		l1.append(firstFile.min_cost())

		l2.append(secondFile.min_cost())

		l1 = sorted(l1)

		l2 = sorted(l2)

	

	print("d =", firstFile.k," first file got a median of :", l1[int(len(l1)/2)], "while the second one got a median of: " , l2[int(len(l2)/2)], "		(10 iterations)")

	res.append(l1[int(len(l1)/2)]>l2[int(len(l2)/2)])

	

	firstFile.k, secondFile.k = 1000000, 1000000

	

	l1, l2 = [], []

	

	for i in range(10):

		l1.append(firstFile.min_cost())

		l2.append(secondFile.min_cost())

		l1 = sorted(l1)

		l2 = sorted(l2)

	

	print("d =", firstFile.k," first file got a median of :", l1[int(len(l1)/2)], "while the second one got a median of: " , l2[int(len(l2)/2)], "		(10 iterations)")

	res.append(l1[int(len(l1)/2)]>l2[int(len(l2)/2)])

	

	print("\n=============================================================\n")

	

	print("Function : CLOSEST_WORD_AMONG")

	

	l1, l2 = [], []

	

	for i in range(10):

		k = 10000

		perm1 = R.closest_word_among(m=10)

		perm2 = R.closest_word_among(m=10)

		for i in range(k):

			cost1 = next(perm1)[1]

			cost2 = next(perm2)[1]

		l1.append(cost1)

		l2.append(cost2)

	l1 = sorted(l1)

	l2 = sorted(l2)

	

	print("d =", k," and m = 10, first file got a median of :", l1[int(len(l1)/2)], "while the second one got a median of: " , l2[int(len(l2)/2)], "		(10 iterations)")

	res.append(l1[int(len(l1)/2)]>l2[int(len(l2)/2)])

	

	l1, l2 = [], []

	

	for i in range(10):

		k = 10000

		perm1 = R.closest_word_among(m=100)

		perm2 = R.closest_word_among(m=100)

		for i in range(k):

			cost1 = next(perm1)[1]

			cost2 = next(perm2)[1]

		l1.append(cost1)

		l2.append(cost2)

	l1 = sorted(l1)

	l2 = sorted(l2)

	

	print("d =", k," and m = 100, first file got a median of :", l1[int(len(l1)/2)], "while the second one got a median of: " , l2[int(len(l2)/2)], "		(10 iterations)")

	res.append(l1[int(len(l1)/2)]>l2[int(len(l2)/2)])

	

	l1, l2 = [], []

	

	for i in range(10):

		k = 100000

		perm1 = R.closest_word_among(m=10)

		perm2 = R.closest_word_among(m=10)

		for i in range(k):

			cost1 = next(perm1)[1]

			cost2 = next(perm2)[1]

		l1.append(cost1)

		l2.append(cost2)

	l1 = sorted(l1)

	l2 = sorted(l2)

	

	print("d =", k," and m = 10, first file got a median of :", l1[int(len(l1)/2)], "while the second one got a median of: " , l2[int(len(l2)/2)], "		(10 iterations)")

	res.append(l1[int(len(l1)/2)]>l2[int(len(l2)/2)])

	

	l1, l2 = [], []

	

	for i in range(10):

		k = 100000

		perm1 = R.closest_word_among(m=100)

		perm2 = R.closest_word_among(m=100)

		for i in range(k):

			cost1 = next(perm1)[1]

			cost2 = next(perm2)[1]

		l1.append(cost1)

		l2.append(cost2)

	l1 = sorted(l1)

	l2 = sorted(l2)

	

	print("d =", k," and m = 100, first file got a median of :", l1[int(len(l1)/2)], "while the second one got a median of: " , l2[int(len(l2)/2)], "		(10 iterations)")

	res.append(l1[int(len(l1)/2)]>l2[int(len(l2)/2)])

	

	print("\n=============================================================\n")

	

	print("Function : K-MEANS")

	

	l1, l2 = [], []

	

	perm1 = firstFile.random_permutation()

	perm2 = secondFile.random_permutation()

	for i in range(mini_D):

		l1.append(next(perm1))

		l2.append(next(perm2))

	

	cost1 = Robustness(l1).k_means(algo = 1)

	cost2 = Robustness(l2).k_means(algo = 1)

	print(x[1])

	

	print("d =", mini_D," and k = 10 with cost_rate, first file got a cost of :", cost1[1], "while the second one got a cost of: " , cost2[1])

	res.append(cost1[1]>l2cost2[2])

	

	print("\n=============================================================\n")

	

	print("CONCLUSION")

	

	res = sorted(res)

	

	if l[int(len(l)/2)] == True:

		print("On average, the first file is more robust than the second file.")

	else:

		print("On average, the second file is more robust than the first file.")