{"id":932365,"date":"2024-12-26T17:56:48","date_gmt":"2024-12-26T09:56:48","guid":{"rendered":"https:\/\/docs.pingcode.com\/ask\/ask-ask\/932365.html"},"modified":"2024-12-26T17:56:51","modified_gmt":"2024-12-26T09:56:51","slug":"python-%e5%a6%82%e4%bd%95%e4%bf%9d%e5%ad%98%e7%89%b9%e5%be%81","status":"publish","type":"post","link":"https:\/\/docs.pingcode.com\/ask\/932365.html","title":{"rendered":"python \u5982\u4f55\u4fdd\u5b58\u7279\u5f81"},"content":{"rendered":"

\"python<\/p>\n

Python\u4fdd\u5b58\u7279\u5f81\u7684\u65b9\u6cd5\u5305\u62ec\uff1a\u4f7f\u7528pickle\u6a21\u5757\u4fdd\u5b58\u4e3a\u6587\u4ef6\u3001\u4f7f\u7528joblib\u6a21\u5757\u8fdb\u884c\u538b\u7f29\u5b58\u50a8\u3001\u5c06\u7279\u5f81\u5b58\u5165\u6570\u636e\u5e93\u3001\u4fdd\u5b58\u4e3aCSV\u6216Excel\u6587\u4ef6\u3001\u5c06\u7279\u5f81\u4fdd\u5b58\u4e3aNumPy\u6570\u7ec4\u7b49\u3002<\/strong>\u5176\u4e2d\uff0c\u4f7f\u7528pickle\u6a21\u5757\u662f\u6700\u5e38\u89c1\u7684\u65b9\u6cd5\uff0c\u56e0\u4e3a\u5b83\u80fd\u591f\u5e8f\u5217\u5316\u51e0\u4e4e\u6240\u6709Python\u5bf9\u8c61\uff0c\u5e76\u4e14\u4f7f\u7528\u7b80\u5355\u65b9\u4fbf\u3002\u4e0b\u9762\u5c06\u8be6\u7ec6\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528pickle\u6a21\u5757\u4fdd\u5b58\u7279\u5f81\u3002<\/p>\n<\/p>\n

\u4e00\u3001\u4f7f\u7528PICKLE\u6a21\u5757<\/h3>\n<\/p>\n

1\u3001\u4ec0\u4e48\u662fPickle\u6a21\u5757<\/h4>\n<\/p>\n

Pickle\u662fPython\u7684\u4e00\u4e2a\u5185\u7f6e\u6a21\u5757\uff0c\u63d0\u4f9b\u4e86\u4e00\u4e2a\u7b80\u5355\u7684\u65b9\u5f0f\u6765\u5e8f\u5217\u5316\u548c\u53cd\u5e8f\u5217\u5316Python\u5bf9\u8c61\u3002\u5e8f\u5217\u5316\u662f\u6307\u5c06Python\u5bf9\u8c61\u8f6c\u6362\u4e3a\u4e00\u4e2a\u5b57\u8282\u6d41\uff0c\u4ee5\u4fbf\u80fd\u591f\u4fdd\u5b58\u5230\u6587\u4ef6\u4e2d\u6216\u901a\u8fc7\u7f51\u7edc\u8fdb\u884c\u4f20\u8f93\u3002\u53cd\u5e8f\u5217\u5316\u5219\u662f\u5c06\u5b57\u8282\u6d41\u91cd\u65b0\u8f6c\u6362\u4e3aPython\u5bf9\u8c61\u7684\u8fc7\u7a0b\u3002<\/p>\n<\/p>\n

2\u3001\u4f7f\u7528Pickle\u4fdd\u5b58\u7279\u5f81<\/h4>\n<\/p>\n

\u8981\u4f7f\u7528Pickle\u4fdd\u5b58\u7279\u5f81\uff0c\u6211\u4eec\u9996\u5148\u9700\u8981\u5c06\u7279\u5f81\u6570\u636e\u8f6c\u6362\u4e3aPython\u5bf9\u8c61\uff08\u5982\u5217\u8868\u3001\u5b57\u5178\u3001NumPy\u6570\u7ec4\u7b49\uff09\uff0c\u7136\u540e\u4f7f\u7528Pickle\u7684dump<\/code>\u51fd\u6570\u5c06\u5bf9\u8c61\u4fdd\u5b58\u5230\u6587\u4ef6\u4e2d\u3002\u4ee5\u4e0b\u662f\u4e00\u4e2a\u7b80\u5355\u7684\u793a\u4f8b\uff1a<\/p>\n<\/p>\n

import pickle<\/p>\n
\u5047\u8bbe\u6709\u4e00\u4e2a\u7279\u5f81\u5b57\u5178<\/strong><\/h2>\n
features = {'feature1': [1, 2, 3], 'feature2': [4, 5, 6]}<\/p>\n
\u5c06\u7279\u5f81\u4fdd\u5b58\u5230\u6587\u4ef6<\/strong><\/h2>\n
with open('features.pkl', 'wb') as f:<\/p>\n
    pickle.dump(features, f)<\/p>\n
<\/code><\/pre>\n<\/p>\n
3\u3001\u8bfb\u53d6Pickle\u6587\u4ef6\u4e2d\u7684\u7279\u5f81<\/h4>\n<\/p>\n
\u8981\u8bfb\u53d6Pickle\u6587\u4ef6\u4e2d\u7684\u7279\u5f81\uff0c\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528Pickle\u7684load<\/code>\u51fd\u6570\uff1a<\/p>\n<\/p>\n
with open('features.pkl', 'rb') as f:<\/p>\n
    loaded_features = pickle.load(f)<\/p>\n
print(loaded_features)<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e8c\u3001\u4f7f\u7528JOBLIB\u6a21\u5757<\/h3>\n<\/p>\n
1\u3001\u4ec0\u4e48\u662fJoblib\u6a21\u5757<\/h4>\n<\/p>\n
Joblib\u662f\u4e00\u4e2a\u7528\u4e8ePython\u7684\u8f7b\u91cf\u7ea7\u5e76\u884c\u8ba1\u7b97\u548c\u6301\u4e45\u5316\u5e93\u3002\u4e0ePickle\u76f8\u6bd4\uff0cJoblib\u66f4\u9002\u5408\u7528\u4e8e\u5b58\u50a8\u5927\u578b\u6570\u636e\uff0c\u56e0\u4e3a\u5b83\u80fd\u591f\u9ad8\u6548\u5730\u538b\u7f29\u548c\u5b58\u50a8\u5927\u89c4\u6a21\u7684NumPy\u6570\u7ec4\u3002<\/p>\n<\/p>\n
2\u3001\u4f7f\u7528Joblib\u4fdd\u5b58\u7279\u5f81<\/h4>\n<\/p>\n
\u4f7f\u7528Joblib\u4fdd\u5b58\u7279\u5f81\u975e\u5e38\u7b80\u5355\uff0c\u53ea\u9700\u4f7f\u7528joblib.dump<\/code>\u51fd\u6570\uff1a<\/p>\n<\/p>\n
from joblib import dump, load<\/p>\n
\u5047\u8bbe\u6709\u4e00\u4e2a\u7279\u5f81\u5b57\u5178<\/strong><\/h2>\n
features = {'feature1': [1, 2, 3], 'feature2': [4, 5, 6]}<\/p>\n
\u5c06\u7279\u5f81\u4fdd\u5b58\u5230\u6587\u4ef6<\/strong><\/h2>\n
dump(features, 'features.joblib')<\/p>\n
<\/code><\/pre>\n<\/p>\n
3\u3001\u8bfb\u53d6Joblib\u6587\u4ef6\u4e2d\u7684\u7279\u5f81<\/h4>\n<\/p>\n
\u540c\u6837\uff0c\u53ef\u4ee5\u4f7f\u7528joblib.load<\/code>\u51fd\u6570\u6765\u8bfb\u53d6\u7279\u5f81\uff1a<\/p>\n<\/p>\n
loaded_features = load('features.joblib')<\/p>\n
print(loaded_features)<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e09\u3001\u4fdd\u5b58\u4e3aCSV\u6216EXCEL\u6587\u4ef6<\/h3>\n<\/p>\n
1\u3001\u4ec0\u4e48\u662fCSV\u548cExcel\u6587\u4ef6<\/h4>\n<\/p>\n
CSV\uff08Comma-Separated Values\uff09\u548cExcel\u6587\u4ef6\u90fd\u662f\u5e38\u7528\u7684\u6570\u636e\u5b58\u50a8\u683c\u5f0f\u3002CSV\u6587\u4ef6\u662f\u4e00\u79cd\u7eaf\u6587\u672c\u683c\u5f0f\uff0c\u7528\u4e8e\u5b58\u50a8\u8868\u683c\u6570\u636e\uff0c\u800cExcel\u6587\u4ef6\u5219\u662fMicrosoft Excel\u4f7f\u7528\u7684\u4e13\u6709\u683c\u5f0f\u3002<\/p>\n<\/p>\n
2\u3001\u4f7f\u7528Pandas\u4fdd\u5b58\u7279\u5f81\u4e3aCSV\u6587\u4ef6<\/h4>\n<\/p>\n
Pandas\u662f\u4e00\u4e2a\u5f3a\u5927\u7684\u6570\u636e\u5904\u7406\u5e93\uff0c\u63d0\u4f9b\u4e86\u5c06DataFrame\u4fdd\u5b58\u4e3aCSV\u6587\u4ef6\u7684\u529f\u80fd\uff1a<\/p>\n<\/p>\n
import pandas as pd<\/p>\n
\u5047\u8bbe\u6709\u4e00\u4e2a\u7279\u5f81DataFrame<\/strong><\/h2>\n
features_df = pd.DataFrame({'feature1': [1, 2, 3], 'feature2': [4, 5, 6]})<\/p>\n
\u5c06\u7279\u5f81\u4fdd\u5b58\u4e3aCSV\u6587\u4ef6<\/strong><\/h2>\n
features_df.to_csv('features.csv', index=False)<\/p>\n
<\/code><\/pre>\n<\/p>\n
3\u3001\u4f7f\u7528Pandas\u4fdd\u5b58\u7279\u5f81\u4e3aExcel\u6587\u4ef6<\/h4>\n<\/p>\n
Pandas\u540c\u6837\u652f\u6301\u5c06DataFrame\u4fdd\u5b58\u4e3aExcel\u6587\u4ef6\uff1a<\/p>\n<\/p>\n
# \u5c06\u7279\u5f81\u4fdd\u5b58\u4e3aExcel\u6587\u4ef6<\/p>\n
features_df.to_excel('features.xlsx', index=False)<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u56db\u3001\u5c06\u7279\u5f81\u5b58\u5165\u6570\u636e\u5e93<\/h3>\n<\/p>\n
1\u3001\u5c06\u7279\u5f81\u5b58\u5165SQL\u6570\u636e\u5e93<\/h4>\n<\/p>\n
Python\u63d0\u4f9b\u4e86\u591a\u79cd\u5e93\u7528\u4e8e\u4e0eSQL\u6570\u636e\u5e93\u4ea4\u4e92\uff0c\u5982SQLite\u3001MySQL\u3001PostgreSQL\u7b49\u3002\u4ee5\u4e0b\u662f\u4f7f\u7528SQLite\u5c06\u7279\u5f81\u5b58\u5165\u6570\u636e\u5e93\u7684\u793a\u4f8b\uff1a<\/p>\n<\/p>\n
import sqlite3<\/p>\n
\u521b\u5efa\u6570\u636e\u5e93\u8fde\u63a5<\/strong><\/h2>\n
conn = sqlite3.connect('features.db')<\/p>\n
c = conn.cursor()<\/p>\n
\u521b\u5efa\u7279\u5f81\u8868<\/strong><\/h2>\n
c.execute('''CREATE TABLE IF NOT EXISTS features<\/p>\n
             (feature1 INTEGER, feature2 INTEGER)''')<\/p>\n
\u63d2\u5165\u7279\u5f81\u6570\u636e<\/strong><\/h2>\n
features = [(1, 4), (2, 5), (3, 6)]<\/p>\n
c.executemany('INSERT INTO features VALUES (?, ?)', features)<\/p>\n
\u63d0\u4ea4\u5e76\u5173\u95ed\u8fde\u63a5<\/strong><\/h2>\n
conn.commit()<\/p>\n
conn.close()<\/p>\n
<\/code><\/pre>\n<\/p>\n
2\u3001\u5c06\u7279\u5f81\u5b58\u5165NoSQL\u6570\u636e\u5e93<\/h4>\n<\/p>\n
NoSQL\u6570\u636e\u5e93\u5982MongoDB\u3001Cassandra\u7b49\u4e5f\u5e7f\u6cdb\u7528\u4e8e\u5b58\u50a8\u5927\u89c4\u6a21\u6570\u636e\u3002\u4ee5\u4e0b\u662f\u5c06\u7279\u5f81\u5b58\u5165MongoDB\u7684\u793a\u4f8b\uff1a<\/p>\n<\/p>\n
from pymongo import MongoClient<\/p>\n
\u521b\u5efaMongoDB\u8fde\u63a5<\/strong><\/h2>\n
client = MongoClient('localhost', 27017)<\/p>\n
db = client['feature_db']<\/p>\n
collection = db['features']<\/p>\n
\u63d2\u5165\u7279\u5f81\u6570\u636e<\/strong><\/h2>\n
features = [{'feature1': 1, 'feature2': 4},<\/p>\n
            {'feature1': 2, 'feature2': 5},<\/p>\n
            {'feature1': 3, 'feature2': 6}]<\/p>\n
collection.insert_many(features)<\/p>\n
\u5173\u95ed\u8fde\u63a5<\/strong><\/h2>\n
client.close()<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e94\u3001\u4fdd\u5b58\u4e3aNUMPY\u6570\u7ec4<\/h3>\n<\/p>\n
1\u3001\u4ec0\u4e48\u662fNumPy\u6570\u7ec4<\/h4>\n<\/p>\n
NumPy\u662fPython\u7684\u4e00\u4e2a\u79d1\u5b66\u8ba1\u7b97\u5e93\uff0c\u63d0\u4f9b\u4e86\u9ad8\u6548\u7684\u591a\u7ef4\u6570\u7ec4\u5bf9\u8c61\u2014\u2014ndarray\u3002NumPy\u6570\u7ec4\u662f\u4e00\u79cd\u5185\u5b58\u8282\u7701\u3001\u9ad8\u6548\u7684\u5b58\u50a8\u65b9\u5f0f\uff0c\u9002\u5408\u7528\u4e8e\u5927\u89c4\u6a21\u6570\u503c\u8ba1\u7b97\u3002<\/p>\n<\/p>\n
2\u3001\u4f7f\u7528NumPy\u4fdd\u5b58\u7279\u5f81<\/h4>\n<\/p>\n
NumPy\u63d0\u4f9b\u4e86\u5c06\u6570\u7ec4\u4fdd\u5b58\u4e3a\u4e8c\u8fdb\u5236\u6587\u4ef6\u548c\u6587\u672c\u6587\u4ef6\u7684\u529f\u80fd\uff1a<\/p>\n<\/p>\n
import numpy as np<\/p>\n
\u5047\u8bbe\u6709\u4e00\u4e2a\u7279\u5f81\u6570\u7ec4<\/strong><\/h2>\n
features = np.array([[1, 4], [2, 5], [3, 6]])<\/p>\n
\u5c06\u7279\u5f81\u4fdd\u5b58\u4e3a\u4e8c\u8fdb\u5236\u6587\u4ef6<\/strong><\/h2>\n
np.save('features.npy', features)<\/p>\n
\u5c06\u7279\u5f81\u4fdd\u5b58\u4e3a\u6587\u672c\u6587\u4ef6<\/strong><\/h2>\n
np.savetxt('features.txt', features, fmt='%d')<\/p>\n
<\/code><\/pre>\n<\/p>\n
3\u3001\u8bfb\u53d6NumPy\u6587\u4ef6\u4e2d\u7684\u7279\u5f81<\/h4>\n<\/p>\n
\u53ef\u4ee5\u4f7f\u7528np.load<\/code>\u548cnp.loadtxt<\/code>\u51fd\u6570\u6765\u8bfb\u53d6\u7279\u5f81\uff1a<\/p>\n<\/p>\n
# \u8bfb\u53d6\u4e8c\u8fdb\u5236\u6587\u4ef6<\/p>\n
loaded_features = np.load('features.npy')<\/p>\n
print(loaded_features)<\/p>\n
\u8bfb\u53d6\u6587\u672c\u6587\u4ef6<\/strong><\/h2>\n
loaded_features_txt = np.loadtxt('features.txt', dtype=int)<\/p>\n
print(loaded_features_txt)<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u516d\u3001\u9009\u62e9\u5408\u9002\u7684\u7279\u5f81\u4fdd\u5b58\u65b9\u6cd5<\/h3>\n<\/p>\n
\u5728\u9009\u62e9\u7279\u5f81\u4fdd\u5b58\u65b9\u6cd5\u65f6\uff0c\u9700\u8981\u6839\u636e\u5177\u4f53\u7684\u5e94\u7528\u573a\u666f\u8fdb\u884c\u9009\u62e9\uff1a<\/p>\n<\/p>\n