动手实现Apriori算法
def createC1(dataset):
C1 = set()
for t in dataset:
for item in t:
item_set = frozenset([item])
C1.add(item_set)
return C1
def scanD(D, ck, minsupport):
ssCnt = {}
for tid in D:
for can in ck:
if can.issubset(tid):
if can not in ssCnt.keys():
ssCnt[can] = 1
else:
ssCnt[can] += 1
numItems = len(D)
reList = []
supportData = {}
for key in ssCnt:
support = ssCnt[key]/numItems
if support >= minsupport:
reList.insert(0, key)
supportData[key] = support
return reList, supportData
def aprioriGen(Lk, k):
retList = []
lenLk = len(Lk)
for i in range(lenLk):
for j in range(i+1, lenLk):
L1 = list(Lk[i])[k:-2]
L2 = list(Lk[j])[:k-2]
if L1 == L2:
retList.append(Lk[i] | Lk[j])
return retList
def apriori(dataSet, minsupport=0.5):
‘’’
apriori算法
:param dataSet: 原始数据集
:param minsupport: 最小支持度阈值
:return: 候选频繁项集列表,所有候选频繁项集的支持度(包括低于阈值的),类型为 (list, dict)
‘’’
#Begin#
C1 = createC1(dataSet)
D = list(map(set, dataSet))
L1, supportData = scanD(D, C1, minsupport)
L =[L1]
k=2
while (len(L[k-2])>0):
Ck = aprioriGen(L[k - 2], k)
Lk, supK = scanD(D, Ck, minsupport)
supportData.update(supK)
L.append(Lk)
k += 1
return L, supportData
#End#
从频繁项集中挖掘关联规则
from utils import apriori, aprioriGen
def calcConf(freqSet, H, supportData, brl, minConf = 0.7):
prunedH = []
for conseq in H:
conf = supportData[freqSet]/supportData[freqSet - conseq]
if conf >= minConf and conf!=1:
brl.append((freqSet - conseq, conseq, conf))
prunedH.append(conseq)
return prunedH
def ruleFromConseq(freqSet, H, supportData, brl, minConf = 0.7):
m = len(H[0])
if len(freqSet) > m+1:
Hmp1 = aprioriGen(H, m+1)
Hmp1 = calcConf(freqSet, Hmp1, supporData, brl, minConf)
if len(Hmp1) > 1:
ruleFromConseq(freqSet, Hmp1, supportData, brl, minConf)
def getSubset(fromList, toList):
for i in range(len(fromList)):
t = [fromList[i]]
tt = frozenset(set(fromList) - set(t))
if not tt in toList:
toList.append(tt)
tt = list(tt)
if len(tt) > 1:
getSubset(tt, toList)
def generateRules(dataset, minsupport, minConf):
‘’’
生成关联规则,可以使用apriori函数获得数据集中的频繁项集列表与支持度
:param dataset:数据集,类型为list
:param minsupport:最小支持度,类型为float
:param minConf:最小可信度,类型为float
:return:关联规则列表,类型为list
‘’’
#Begin#
L, supportData = apriori(dataset, minsupport)
bigRuleList = []
for i in range(1, len(L)):
for freqSet in L[i]:
H1 = [] #对每个频繁项集构了,建只包含单个元素的集合,即可以出现在规则右边
for item in freqSet:
H1.append(frozenset([item]))
if (i > 1):
H1 = calcConf(freqSet,H1,supportData,bigRuleList,minConf)
#包含三个及以上元素的频繁集
rulesFromConseq(freqSet,H1,supportData,bigRuleList,minConf)
else:
#包含两个元素的频繁集
calcConf(freqSet,H1,supportData,bigRuleList,minConf)
return bigRuleList
#End#
超市购物清单关联规则分析
from utils import generateRules
import pandas as pd
dic={‘yogurt’:1, ‘pork’:2, ‘sandwich bags’:3, ‘lunch meat’:4, ‘all- purpose’:5, ‘flour’:6, ‘soda’:7, ‘butter’:8, ‘vegetables’:9, ‘beef’:10, ‘aluminum foil’:11, ‘dinner rolls’:12, ‘shampoo’:13, ‘mixes’:14, ‘soap’:15, ‘laundry detergent’:16, ‘ice cream’:17, ‘toilet paper’:18, ‘hand soap’:19, ‘waffles’:20, ‘cheeses’:21, ‘milk’:22, ‘dishwashing liquid/detergent’:23, ‘individual meals’:24, ‘cereals’:25, ‘tortillas’:26, ‘spaghetti sauce’:27, ‘ketchup’:28, ‘sandwich loaves’:29, ‘poultry’:30, ‘bagels’:31, ‘eggs’:32, ‘juice’:33, ‘pasta’:34, ‘paper towels’:35, ‘coffee/tea’:36, ‘fruits’:37, ‘sugar’:38}
def genRules(data_path, min_support, min_conf):
‘’’
对文件路径为data_path的数据进行关联规则挖掘
:param data_path: 数据集路径
:param min_support: 最小支持度
:param min_conf: 最小置信度
:return:关联规则信息,类型为list
‘’’
# Begin#
data = pd.read_csv(data_path)
data[‘good’] = data[‘good’].apply(lambda x:’,’+x)
data = data.groupby(‘id’).sum().reset_index()
data[‘good’] = data[‘good’].apply(lambda x :[x[1:]])
data_list = list(data[‘good’])
data_translation = []
for i in data_list:
p = i[0].split(’,’)
a = []
for j in p:
a.append(dic.get(j))
data_translation.append(a)
rule = generateRules(data_translation,min_support, min_conf)
return rule
#End#