Implementation of KNN and Naive_Bayes

# 导入数据集
from sklearn import datasets
from sklearn.model_selection import train_test_split
from sklearn import neighbors

datasets = datasets.load_breast_cancer()
X = datasets.data;
y = datasets.target;
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.3, random_state = 0)
k = 5
# print(datasets.DESCR)
# malignant - 0, benign - 1
y_predict = []

def knn(X_train, y_train, X_test, y_predict):
    '''
    对测试集的数据进行预测，得到的结果与y_test比较。用欧式距离进行计算。
    '''
    for test_data in X_test:
        first_k_instance = []
        for i in range(len(X_train)):
            distance = 0;
            for attributes_no in range(len(X_train[0])):
                distance += (test_data[attributes_no] - X_train[i][attributes_no]) ** 2
            Euclid_distance = distance ** 0.5
            #print(Euclid_distance)
            if i < k:
                first_k_instance.append((i, Euclid_distance))
            elif Euclid_distance < first_k_instance[k-1][1]:
                first_k_instance[k-1] = (i, Euclid_distance)
            first_k_instance = sorted(first_k_instance, key = lambda x:x[1]) 
            #print(first_k_instance)
        # 现在得到了距离测试点最近的k个点，用多数表决器来判断测试点是良性还是恶性
        benign = 0
        malignant = 0
        for instance in first_k_instance:
            if y_train[instance[0]] == 0:
                malignant += 1
            else:
                benign += 1
        if malignant >= benign:
            y_predict.append(0)
        else:
            y_predict.append(1)

def accuracy(y_predict, y_test):
    correct = 0
    for i in range(len(y_predict)):
        if y_predict[i] == y_test[i]:
            correct += 1
    accuracy_rate = correct / len(y_predict)
    return correct, accuracy_rate

def main():
    knn(X_train, y_train, X_test, y_predict)
    correct, accuracy_rate = accuracy(y_predict, y_test)
    print(y_predict)
    print("kNN模型测试集预测的准确率为：%.3f" % accuracy_rate);
    KNN = neighbors.KNeighborsClassifier(n_neighbors = 5)
    KNN.fit(X_train, y_train)
    print("sklearn库中kNN模型预测的准确率为：%.3f" % KNN.score(X_test, y_test));
    
if __name__ == '__main__':
    main()

# load datasets
from sklearn import datasets
from sklearn.model_selection import train_test_split
from sklearn import naive_bayes

datasets = datasets.load_breast_cancer()
X = datasets.data;
y = datasets.target;
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.3, random_state = 0)
#print(datasets.DESCR)
#malignant - 0, benign - 1
y_predict = []

def distribution(X_train, y_train):
    '''
    先把区间分好，然后再计算概率。
    '''
    #===============区间划分====================#
    attributes_max_min_mean = []
    # 记录所有属性的最大值、最小值和平均值
    for i in range(len(X_train[0])):
        #属性循环
        #section = [max, min, mean]
        section = [X_train[0][i], X_train[0][i], 0]
        for instance in X_train:
            #训练样例循环
            if instance[i] > section[0]:
                section[0] = instance[i]
            if instance[i] < section[1]:
                section[1] = instance[i]
            section[2] += instance[i]
        section[2] /= len(X_train)
        attributes_max_min_mean.append(section)
        
    #=========计算每个属性落在每个区间的样例个数=========#
    instance_distribution = []
    for i in range(len(X_train[0])):
        #属性循环
        smaller_benign = 0
        larger_benign = 0
        smaller_malignant = 0
        larger_malignant = 0
        for j in range(len(X_train)):
            #训练样例循环
            if X_train[j][i] > attributes_max_min_mean[i][2]:
                if y_train[j] == 1:
                    larger_benign += 1
                else:
                    larger_malignant +=1
            elif y_train[j] == 1:
                smaller_benign += 1
            else:
                smaller_malignant += 1   
        instance_distribution.append([smaller_benign, larger_benign, smaller_malignant, larger_malignant])
        
    return instance_distribution, attributes_max_min_mean

def Naive_Bayes(X_test, y_predict, instance_distribution,attributes_max_min_mean):
    for test_data in X_test:
        #测试样例循环
        #训练集中良性和恶性肿瘤的数量
        malignant = instance_distribution[0][2] + instance_distribution[0][3]
        benign = instance_distribution[0][0] + instance_distribution[0][1]
        #概率初始化，下面计算每个属性的概率
        p_xc0 = 1
        p_xc1 = 1
        for i in range(len(X_train[0])):
            # 属性循环
            if test_data[i] > attributes_max_min_mean[i][2]:
                p_xc0 *= instance_distribution[i][3] / malignant
                p_xc1 *= instance_distribution[i][1] / benign
            else:
                p_xc0 *= instance_distribution[i][2] / malignant
                p_xc1 *= instance_distribution[i][0] / benign
        p0 = p_xc0 * malignant / (malignant + benign)
        p1 = p_xc1 * benign / (malignant + benign)
        if p0 > p1:
            y_predict.append(0)
        else:
            y_predict.append(1)      

def accuracy(y_predict, y_test):
    correct = 0
    for i in range(len(y_predict)):
        if y_predict[i] == y_test[i]:
            correct += 1
    accuracy_rate = correct / len(y_predict)
    return correct, accuracy_rate

def main():
    instance_distribution, attributes_max_min_mean = distribution(X_train, y_train)
    Naive_Bayes(X_test, y_predict, instance_distribution, attributes_max_min_mean)
    correct, accuracy_rate = accuracy(y_predict, y_test)
    print(y_predict)
    print("Naive Bayes模型测试集预测的准确率为：%.3f" % accuracy_rate);
    bayes = naive_bayes.GaussianNB()
    bayes.fit(X_train, y_train)
    print("sklearn库中Naive Bayes模型预测的准确率为：%.3f" % bayes.score(X_test, y_test));
    
if __name__ == '__main__':
    main()