数据挖掘-Apriori算法

from collections import defaultdict
import pandas as pd

def load_data(file):
    res=[]
    with open(file) as f:
        for i in f:
            res.append(list(map(int,i.split())))
    return res
def tel_k_(a,b,k):
    a.sort()
    b.sort()
    if k==0:
        return True
    for i in range(k):
        if a[i]!=b[i]:
            return False
    return True

def generate_c(l):
    l.sort()
    n=len(l)
    C = defaultdict(int)
    if n==0:
        return C
    k=len(l[0])-1

    for i in range(n-1):
        for j in range(i+1,n):
            if tel_k_(l[i],l[j],k):
                #print(set(l[i]+l[j][k:]))
                C[tuple(l[i]+l[j][k:])]+=0
    return C


def apriori(data,sup):
    threshold=len(data)*sup

    #生成写入文件
    with open("result.txt", "a+") as f:
        f.write("\nminsupport="+str(sup))  # 自带文件关闭功能，不需要再写f.close()

    #生成候选集-1
    c1=defaultdict(int)
    col=[]
    for i in data:
        for x in i:
            if c1[x]==0:
                col.append(x)
            c1[x]+=1
    #print(c1)

    #生成dataframe
    col.sort()
    tmp={}
    for i in col:
        tmp[i]=[0]*len(data)

    dt = pd.DataFrame(tmp)
    for i in range(len(data)):
        for x in data[i]:
            dt.loc[i][x]=1
    print(dt)


    #生成频繁项集-1
    L1=[]
    for i in c1:
        if c1[i]>threshold:
            L1.append([i])
    with open("result.txt", "a+") as f:
        f.write("\nL1:"+str(L1))  # 自带文件关闭功能，不需要再写f.close()
    #生成候选集-2
    C=generate_c(L1)
    k=2
    while C:
        for i in range(len(dt)):
            for j in C:
                #print(i,j)
                flag=True
                for x in j:
                    if dt.loc[i][x]==0:
                        flag=False
                if flag:
                    C[j]+=1
        L=[]
        for i in C:
            if C[i]>threshold:
                L.append(list(i))
        if L:
            with open("result.txt", "a+") as f:
                f.write("\nL"+str(k)+":"+str(L))  # 自带文件关闭功能，不需要再写f.close()
            k+=1
            C=generate_c(L)
        else:
            break
    with open("result.txt", "a+") as f:
        f.write("\n")  # 自带文件关闭功能，不需要再写f.close()
    return

if __name__=='__main__':
    data=load_data('data.txt')
    n=len(data)
    #print(n,data)
    apriori(data,0.22)

运行效果：