Re:■ 나이브 베이즈 알고리즘을 파이썬으로 구현하세요 ~ 답글로 검사받겠습니다.

[10기_김재훈]

class Naive_Bayes:
    def __init__(self, path = 'd:\\datadata\\flu.csv'):
        import csv
       
        a1 = []
        with open(path,'r') as f:
            file = csv.reader(f)
            for i in file:       
                a1.append(i)
       
        len_a1 = len(a1)
       
        Ys = []
        for i in a1:
            Ys.append(i[-1])
        Ys = set(Ys[1:])
       
       
       
        # 딕셔너리 T - 속에 Y 카테고리에 따라 딕셔너리로 열 별 개수 넣을 예정
        T = {}
        for i in Ys :
            T[i] = []
       
        for _ in range(len(a1[0])):
            for key in T.keys():
                T[key].append({})

        # 열 별 개수 넣기
       
        for i in range(1,len(a1)):          # 행       
            for key,value in T.items():
                if a1[i][-1] == key:
                    for j in range(1,len(a1[0])):          # 열
                        aij = a1[i][j]           
                        if aij not in T[key][j]:
                            T[key][j][aij] = 1
                        else :
                            T[key][j][aij] += 1 
                    

        # 개수 -> 확률
          
        for key,value in T.items():
           
            for n,i in enumerate(T[key]):
                #if
                if n < len(a1[0]) -1 :
                    tmp = 0
                    for v in i.values():
                        tmp += v
                    for k,v in i.items():
                        i[k] /= tmp
                else:
                    i[key] /= (len(a1)-1)       
           
        self.T = T
        self.a1 = a1
       
           
           
    def get_result(self,test = [1, 'Y','N','Mild','N','_'] ):
       
        # 평가 데이터 넣어서 계산
       
        # p(x1,x2,....xn | Ck)
        result = [ [], [], [] ]
        for key, value in self.T.items():
            result[0].append(key)           
            tmp = 1
            for i in range(1, len(self.a1[0]) -1):               
                tmp *= self.T[key][i][test[i]]
            result[1].append(tmp)           
       
        # p(Ck) * p(x1,x2,....xn | Ck)
        for i,j in enumerate(result[0]):
            result[1][i] *= self.T[j][len(a1[0])-1][j]       
       
        # 판정
        tmp = 0
        for i in range(len(result[1])):
            if result[1][i]  > tmp :
                tmp = i
       
        return result[0][i]

nb = Naive_Bayes()
print('판단 결과 : ', nb.get_result())