Big-data/dic.py at master · ruchi09/Big-data · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
import random
from collections import namedtuple
import numpy as np

n=5
t=5
m=2
s=2
support = 2
confidence = 0.5
suspect=n
start=0
c = n
base = [0 for j in range(0,n)]
a=[[0 for i in range(t)] for j in range(n)]
it = list()
rules = list()
iteminfo = namedtuple("iteminfo","item count stage dones length")
allrules = namedtuple("allrules","lhs rhs confidence")

def gen_one_size_itemset() :

	for i in range(0,n) :
		temp = iteminfo(item = [0 for j in range(0,n)] ,count = 0 ,stage = 1 ,dones = 0 ,length = 1 )
		it.append(temp)
		it[i].item[i] = 1

def read_m_trans() :

	global start
	for i in range (start*m,start*m+m):
		for j in range (0,c):
			if (list(np.array(a[i]) & np.array(it[j].item)) == it[j].item and it[j].dones<t) :
				it[j] = it[j]._replace(count=int(it[j].count)+1)
			it[j] = it[j]._replace(dones=int(it[j].dones)+1)
	start += 1
	if (start == (t/m)) :
		for i in range (start*m,start*m+t%m) :
			for j in range (0,c):
				if (list(np.array(a[i]) & np.array(it[j].item)) == it[j].item and it[j].dones<t) :
					it[j] = it[j]._replace(count = int(it[j].count)+1)
				it[j] = it[j]._replace(dones = int(it[j].dones)+1)
		start = 0

def update_itemset_info() :

	global suspect
	for j in range (0,c):
		if it[j].dones == t :
			suspect -= 1
			if it[j].count >= support :
				it[j] = it[j]._replace(stage=3)
			else :
				it[j] = it[j]._replace(stage=0)

def dic() :

	global start,c,suspect
	while (start < t/m and suspect != 0) :

		read_m_trans()
		update_itemset_info()
		for i in range (2,n) :
			counter = 0
			checklist = list()
			for j in range (0,c) :
				if(it[j].length == i-1 and (it[j].stage == 2 or it[j].stage == 3)) :
					counter += 1
			if(counter >= i) :
				for j in range (0,c) :
					if(it[j].length == i-1) :
						for k in range (0,n) :
							if(list(np.array(it[j].item) & np.array(it[k].item)) == base and (it[j].stage == 2 or it[j].stage == 3)) :
								checklist.append(list(np.array(it[j].item) | np.array(it[k].item)))
			temp = 0
			for j in range (0,len(checklist)) :
				if(checklist[j] != base) :
					temp = 1
					for k in range (0, len(checklist)) :
						if( j!= k and checklist[j] == checklist[k]) :
							temp += 1
							checklist[k] = base
				if(temp == i) :
					possible = 1
					for k in range (0,c) :
						if(it[k].item == checklist[j]) :
							possible = 0
							break
					if (possible == 1 and checklist[j] != base) :
						c += 1
						suspect += 1
						temporary = iteminfo(item = checklist[j] ,count = 0,stage = 1,dones = 0,length = i)
						it.append(temporary)
						checklist[j] = base

if __name__ == "__main__":

	a = [[1, 0, 0, 1, 0],[1, 0, 0, 0, 1],[0, 1, 1, 0, 1],[1, 1, 1, 1, 1],[0, 1, 1, 1, 0]]

	# generate one size subsets
	gen_one_size_itemset()

	#start dic
	dic()

	#print the item sets and generated rules
	print "frequent item sets : "
	for i in range (0,c) :
		if(it[i].stage == 3) :
			print it[i]
	for r in range (2,n) :
		lor = r
		for i in range (0,c) :
			for j in range (0,c) :
				if(it[j].length < r and it[i].length == r) :
					if(list(np.array(it[j].item) & np.array(it[i].item)) == it[j].item) :
						for k in range (0,c) :
							if((it[k].length == r-it[j].length)and(list(np.array(it[k].item) | np.array(it[j].item)) == it[i].item)) :
								if ((it[i].count/float(it[j].count)) >= confidence) :
									temp = allrules(lhs = it[j].item ,rhs = it[k].item, confidence = it[i].count/float(it[j].count))
									rules.append(temp)
	print "rules : "
	for i in range (0,len(rules)) :
		print rules[i]