贝叶斯算法python代码 不要用scikit-learn

贝叶斯定理是一种统计推断方法，它在机器学习中常用于概率分类问题。在Python中，我们可以直接编写朴素贝叶斯算法的代码，而不需要依赖Scikit-learn库。以下是基于特征向量和概率计算的简单朴素贝叶斯分类器示例：

import numpy as np

class NaiveBayesClassifier:
    def __init__(self):
        self.classes = None
        self.class_probs = {}
        self.feature_probs = {}

    def fit(self, X, y):
        self.classes = np.unique(y)
        for class_ in self.classes:
            # 计算每个类别的样本数
            count_class = (y == class_).sum()
            
            # 计算每类的概率 P(class)
            self.class_probs[class_] = count_class / len(y)

            # 对于每一特征，计算属于该类别的条件下该特征出现的概率 P(feature|class)
            feature_counts = np.zeros(X.shape[1])
            for i in range(count_class):
                index = np.where(y[i] == class_)[0][0]
                feature_counts += X[index]

            self.feature_probs[class_] = feature_counts / count_class

    def predict(self, X):
        predictions = []
        for x in X:
            posterior_class_probs = {
                class_: self.class_probs[class_] * np.prod(
                    [self.feature_probs[class_][feature] for feature in x if feature != 0]
                )
                for class_ in self.classes
            }
            predictions.append(max(posterior_class_probs, key=posterior_class_probs.get))
        return predictions

# 使用示例
X = ... # 特征数据
y = ... # 标签数据
clf = NaiveBayesClassifier()
clf.fit(X, y)
predictions = clf.predict(X)