现有好评和差评两个文本文件,我需要对这两个中文文本文件作为训练集,使用sklearn实现贝叶斯分类,区分好评和差评,并给出准确率和结果可视化图,给出示例代码

以下是实现贝叶斯分类的示例代码：

import os
import jieba
from sklearn.feature_extraction.text import CountVectorizer
from sklearn.naive_bayes import MultinomialNB
from sklearn.metrics import accuracy_score
import matplotlib.pyplot as plt

# 读取数据
def load_data():
    pos_file = "./pos.txt"
    neg_file = "./neg.txt"
    pos_data = []
    neg_data = []
    with open(pos_file, "r", encoding="utf-8") as f:
        for line in f.readlines():
            pos_data.append(line.strip())
    with open(neg_file, "r", encoding="utf-8") as f:
        for line in f.readlines():
            neg_data.append(line.strip())
    return pos_data, neg_data

# 分词
def seg_text(text):
    words = jieba.cut(text)
    return " ".join(words)

# 特征提取
def extract_feature(train_data):
    vectorizer = CountVectorizer()
    feature = vectorizer.fit_transform(train_data)
    return vectorizer, feature

# 训练模型
def train_model(pos_data, neg_data):
    pos_data = [seg_text(text) for text in pos_data]
    neg_data = [seg_text(text) for text in neg_data]
    train_data = pos_data + neg_data
    label = [1]*len(pos_data) + [0]*len(neg_data)
    vectorizer, feature = extract_feature(train_data)
    clf = MultinomialNB()
    clf.fit(feature, label)
    return clf, vectorizer

# 预测
def predict(clf, vectorizer, test_data):
    test_data = [seg_text(text) for text in test_data]
    test_feature = vectorizer.transform(test_data)
    return clf.predict(test_feature)

# 计算准确率
def calc_accuracy(label, pred):
    acc = accuracy_score(label, pred)
    return acc

# 可视化结果
def plot_result(pos_acc, neg_acc):
    plt.bar(range(2), [pos_acc, neg_acc], tick_label=["Positive", "Negative"])
    plt.ylim(0, 1)
    plt.ylabel("Accuracy")
    plt.show()

if __name__ == "__main__":
    pos_data, neg_data = load_data()
    train_pos_data = pos_data[:800]  # 取前800条数据作为训练集
    train_neg_data = neg_data[:800]
    test_pos_data = pos_data[800:]   # 取后200条数据作为测试集
    test_neg_data = neg_data[800:]
    clf, vectorizer = train_model(train_pos_data, train_neg_data)
    pos_pred = predict(clf, vectorizer, test_pos_data)
    neg_pred = predict(clf, vectorizer, test_neg_data)
    pos_acc = calc_accuracy([1]*len(pos_pred), pos_pred)
    neg_acc = calc_accuracy([0]*len(neg_pred), neg_pred)
    print("Positive Accuracy: {:.2f}%".format(pos_acc*100))
    print("Negative Accuracy: {:.2f}%".format(neg_acc*100))
    plot_result(pos_acc, neg_acc)

在运行上述代码前，请确保已安装好相关的依赖包，并准备好 `pos.txt` 和 `neg.txt` 两个文本文件，文件中每行为一条评价。代码中将前800条数据作为训练集，后200条数据作为测试集，可以根据实际情况进行调整。

运行结果会输出两个分类的准确率，并给出一个柱状图可视化结果。