In [115]:
import json
import datetime
import numpy as np
In [116]:
# 1,3,4
# 2,3,5
# 1,2,3,5
# 2,5
def loadDataSet():
return [[1, 3, 4], [2, 3, 5], [1, 2, 3, 5], [2, 5]]
dataSet = loadDataSet()
In [117]:
dataSet
Out[117]:
In [118]:
# 找出所有dataset中的item集合
# frozenset 和 set 最大的差異在於前者不可變,後者可變,可當作dict的key
def createC1(dataSet):
C1 = []
for transaction in dataSet:
for item in transaction:
if not [item] in C1:
C1.append([item])
C1.sort()
# print(C1)
return map(frozenset, C1)#use frozen set so we
#can use it as a key in a dict
In [119]:
c1 = createC1(dataSet)
In [120]:
c1
Out[120]:
In [121]:
# scan dataset 輸出高於minSupport的item
# dict supportData 儲存 item support value
def scanD(D, Ck, minSupport):
ssCnt = {}
# 算出每一個item總出現次數
for tid in D:
for can in Ck:
if can.issubset(tid):
if not ssCnt.has_key(can): ssCnt[can]=1
else: ssCnt[can] += 1
numItems = float(len(D))
retList = []
supportData = {}
# retList 儲存 >輸出高於minSupport的items集合
for key in ssCnt:
support = ssCnt[key]/numItems
if support >= minSupport:
# retList.insert(0,key) 從首插入
# retList.append(key) 從尾插入
retList.append(key)
supportData[key] = support
return retList, supportData , ssCnt
In [122]:
retList, supportData , ssCnt = scanD(dataSet, c1, 0.5)
In [123]:
# item4低於0.5
retList
Out[123]:
In [124]:
supportData
Out[124]:
In [127]:
# 從小集合聚合成大集合 兩個兩個聚合
def aprioriGen(Lk, k): #creates Ck
retList = []
lenLk = len(Lk)
# 指針移動尋找集合 i,j各別移動
for i in range(lenLk):
for j in range(i+1, lenLk):
L1 = list(Lk[i])[:k-2]; L2 = list(Lk[j])[:k-2]
# 先排序太精妙了 可以避免重覆的 屌啊～
L1.sort(); L2.sort()
if L1==L2: #if first k-2 elements are equal
retList.append(Lk[i] | Lk[j]) #set union
return retList
# 輸入dataset 從小集合聚合到大集合，輸出高於minSupport的item
def apriori(dataSet, minSupport = 0.5):
C1 = createC1(dataSet)
D = map(set, dataSet)
L1, supportData ,ssCnt= scanD(D, C1, minSupport)
L = [L1]
k = 2
while (len(L[k-2]) > 0):
Ck = aprioriGen(L[k-2], k)
Lk, supK ,ssCnt = scanD(D, Ck, minSupport)#scan DB to get Lk
supportData.update(supK)
L.append(Lk)
k += 1
return L, supportData
In [128]:
L,supportData = apriori(dataSet,0.5)
In [129]:
L
Out[129]:
In [131]:
supportData
Out[131]:
In [133]:
# 算出Rule
def generateRules(L, supportData, minConf=0.7): #supportData is a dict coming from scanD
bigRuleList = []
# L[0] 集合只有一個item frozenset([5]) L[1]集合有兩個item frozenset([2, 5])
for i in range(1, len(L)):#only get the sets with two or more items
for freqSet in L[i]:
H1 = [frozenset([item]) for item in freqSet]
if (i > 1):
rulesFromConseq(freqSet, H1, supportData, bigRuleList, minConf)
else:
# 計算 confidence
calcConf(freqSet, H1, supportData, bigRuleList, minConf)
return bigRuleList
def calcConf(freqSet, H, supportData, brl, minConf=0.7):
prunedH = [] #create new list to return
for conseq in H:
# freqSet U conseq / freqSet - conseq 豆奶：freqSet - conseq 青菜：conseq 豆奶U青菜：freqSet
# 買了豆奶和青菜 豆奶 -> 青菜 買了豆奶之後，再買青菜的機率
conf = supportData[freqSet]/supportData[freqSet-conseq] #calc confidence
if conf >= minConf:
print freqSet-conseq,'-->',conseq,'conf:',conf
brl.append((freqSet-conseq, conseq, conf))
prunedH.append(conseq)
return prunedH
def rulesFromConseq(freqSet, H, supportData, brl, minConf=0.7):
# 每一個集合中的物品數
m = len(H[0])
if (len(freqSet) > (m + 1)): #try further merging
Hmp1 = aprioriGen(H, m+1)#create Hm+1 new candidates
Hmp1 = calcConf(freqSet, Hmp1, supportData, brl, minConf)
if (len(Hmp1) > 1): #need at least two sets to merge
rulesFromConseq(freqSet, Hmp1, supportData, brl, minConf)
In [134]:
rule = generateRules(L,supportData,0.7)
In [135]:
rule
Out[135]:
In [136]:
# data source : retail.dat 4.2Mb
In [137]:
datapath = '/Users/wy/Desktop/retail.dat'
dataList=[]
with open (datapath,'r') as f:
lines = f.readlines()
for index, line in enumerate(lines):
line=line.strip()
listLine = map(int,line.split(' '))
dataList.append(listLine)
In [138]:
start = datetime.datetime.now()
L,supportData = apriori(dataList,0.1)
rule = generateRules(L,supportData,0.2)
end = datetime.datetime.now()
runtime = end - start
In [139]:
runtime
Out[139]:
In [152]:
def store(data,path):
with open(path,'w') as f:
f.write(data)
In [153]:
freItem_list = []
for tmp in L:
if len(tmp) !=0:
for item in tmp:
freItem_list.append(list(item))
freItem_json = json.dumps(freItem_list,indent=4)
store(freItem_json,'/Users/wy/Desktop/freItem.json')
In [158]:
freItem_list
Out[158]:
In [155]:
supportData_list = []
for key, value in supportData.iteritems():
s = str(list(key))+':'+str(value)
supportData_list.append(s)
supportData_list.sort()
supportData_json = json.dumps(supportData_list,indent=4)
store(supportData_json,'/Users/wy/Desktop/supportData.json')
In [160]:
# 前10筆
supportData_list[:10]
Out[160]:
In [156]:
strongRule = []
for tmp in rule:
s = str(list(tmp[0])[0])+'-->'+str(list(tmp[1])[0])+'---'+str(tmp[2])
strongRule.append(s)
strongRule_json = json.dumps(strongRule,indent=4)
store(strongRule_json,'/Users/wy/Desktop/strongRule.json')
In [161]:
strongRule
Out[161]:
In [ ]:
# reference Machine Learning in action