#注意：为了方便分析，建议把代码文件（.ipynb）和数据文件(xls,csv,txt等)放到同一个文件夹/目录

数据分析基本操作

#载入分析模块

import pandas as pd

import numpy as np

from sklearn.preprocessing import StandardScaler

from sklearn.cluster import KMeans

from sklearn.ensemble import RandomForestClassifier

from sklearn.neural_network import MLPClassifier

from sklearn.feature_extraction.text import TfidfVectorizer

from sklearn.decomposition import PCA

from mlxtend.frequent_patterns import apriori

from transformers import pipeline

#  一.多格式数据导入

import pandas as pd

    # CSV文件读取

    data1 = pd.read_csv('上海餐饮数据.csv') 

 #查看读取的数据

 data1.head()

    # Excel文件读取

    data2 = pd.read_excel('中医辨证.xlsx')

#查看读取的数据

   data2.head()

    # TXT文件读取

data3= open('news1.txt', encoding = 'utf-8')

#查看读取的数据

content = data3.read()

print(content)  # 输出全部内容

二.常见数据分析方法

# 2. 关联规则分析

def association_analysis(df):

    basket = df.groupby(['order_id','product_name'])['quantity'].sum().unstack()

    basket = basket.fillna(0).applymap(lambda x: 1 if x>0 else 0)

    frequent_items = apriori(basket, min_support=0.05, use_colnames=True)

    return frequent_items.sort_values('support', ascending=False)

# 3. 分类预测

def classification(X_train, y_train):

    clf = RandomForestClassifier()

    clf.fit(X_train, y_train)

    return clf

# 4. 聚类分析

def clustering(X):

    scaler = StandardScaler()

    X_scaled = scaler.fit_transform(X)

    kmeans = KMeans(n_clusters=4)

    clusters = kmeans.fit_predict(X_scaled)

    return clusters

# 5. 神经网络建模

def neural_net(X_train, y_train):

    mlp = MLPClassifier(hidden_layer_sizes=(100,50), max_iter=500)

    mlp.fit(X_train, y_train)

    return mlp

# 6. 表征学习

def representation_learning(texts):

    tfidf = TfidfVectorizer(max_features=1000)

    tfidf_vectors = tfidf.fit_transform(texts)

    pca = PCA(n_components=50)

    reduced_vectors = pca.fit_transform(tfidf_vectors.toarray())

    return reduced_vectors

# 7. 情感分析

def sentiment_analysis(reviews):

    classifier = pipeline("text-classification", model="bert-base-chinese")

    results = classifier(reviews)

    return pd.DataFrame(results)

# 主程序

if __name__ == "__main__":

    # 数据加载

    orders, products, reviews = load_data()

    

    # 关联分析示例

    frequent_items = association_analysis(orders)

    

    # 分类示例（预测客户是否会回购）

    X_class = products[['price','sales']]

    y_class = products['repurchase']

    clf_model = classification(X_class, y_class)

    

    # 聚类示例（客户分群）

    X_cluster = orders.groupby('user_id').agg({'price':'sum','quantity':'count'})

    user_clusters = clustering(X_cluster)

    

    # 神经网络示例

    nn_model = neural_net(X_class, y_class)

    

    # 表征学习示例

    product_vectors = representation_learning(products['description'])

    

    # 情感分析示例

    sentiment_results = sentiment_analysis(reviews[:100])  # 限制数量避免内存问题