什么是提取/转换/加载 (ETL)？
我尝试阅读维基百科文章“提取、转换、加载”，但这只会让我更加困惑......
有人可以解释一下 ETL 是什么，以及它实际上是如何完成的吗？
I've tried reading the Wikipedia article for "extract, transform, load", but that just leaves me more confused...
Can someone explain what ETL is, and how it is actually done?
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（5）
ETL 正在从一个系统获取数据（提取），对其进行修改（转换）并将其加载到另一个系统（加载）。
并且不一定按照这个顺序。您可以拨打电话或 ELT。但可能不是 LTE。 :-)
这是一个包罗万象的名称，适用于从一个系统获取数据并将其移动到另一个系统的任何进程。
ETL is taking data from one system (extract), modifying it (transform) and loading it into another system (load).
And not necessarily in that order. You can TEL, or ELT. Probably not LTE though. :-)
It's a catch-all name for any process that takes data from one system and moves it to another.
ETL 通常用于数据仓库。它不是加载数据仓库的特定实现，它只是用于填充数据仓库的非常高级的算法。
ETL is commonly used for data warehousing. It's not a specific implementation to load a data warehouse, it's just a very high-level algorithm that should be used to populate a data warehouse.
提取是指从一个或多个数据库中取出数据。
转换意味着更改数据，但您需要更改数据以满足您的业务需求。
Load的意思是把它放到目标数据库中。
Extract means to take data out of one or many databases.
Transform means to change the data however you need it changed to suit the needs of your business.
Load means to put it in the target database.
ETL 是提取、转换、加载三种数据库功能的缩写，这三种功能组合成一个工具，用于从一个数据库中提取数据并将其放入另一个数据库中。
提取是从数据库中读取数据的过程。
转换是将提取的数据从先前的形式转换为所需形式的过程，以便可以将其放入另一个数据库中。通过使用规则或查找表或通过将数据与其他数据组合来进行转换。
Load是将数据写入目标数据库的过程。
ETL 用于将数据从一个数据库迁移到另一个数据库，形成数据集市和数据仓库，并将数据库从一种格式或类型转换为另一种格式或类型。
ETL is short for extract, transform, load, three database functions that are combined into one tool to pull data out of one database and place it into another database.
Extract is the process of reading data from a database.
Transform is the process of converting the extracted data from its previous form into the form it needs to be in so that it can be placed into another database. Transformation occurs by using rules or lookup tables or by combining the data with other data.
Load is the process of writing the data into the target database.
ETL is used to migrate data from one database to another, to form data marts and data warehouses and also to convert databases from one format or type to another.
ETL（提取、转换、加载）过程通过促进数据的获取、准备和集成以进行分析和建模，在数据科学中发挥着重要作用。在本文中，我们将在数据科学的背景下深入研究 ETL 流程，检查其关键组件和最佳实践。
提取：
数据科学 ETL 过程的第一步是数据提取。数据可以来自各种位置，包括数据库、API、网络抓取、传感器数据、社交媒体平台等。提取阶段涉及识别相关数据源并检索所需数据。这可能需要查询数据库、发出 API 请求或利用网络抓取技术。提取的数据可以是结构化的、半结构化的或非结构化的，并且可以涵盖文本、数值、图像或其他形式的数据。
转换：
ETL 过程中的转换步骤对于数据科学至关重要。它涉及清理、预处理和操作提取的数据，使其适合分析和建模。此阶段包含数据清理、缺失值插补、数据标准化、特征工程、降维和数据聚合等任务。数据科学家可以在此阶段使用各种技术和算法，具体取决于数据的性质和分析的目标。
加载：
数据科学 ETL 过程的最后一步是数据加载。数据转换后，需要将其加载为合适的格式或结构以供进一步分析。这可能涉及将数据存储在数据库、数据湖或特定文件格式中。在加载过程中确保数据完整性和安全性以及建立适当的数据治理实践以遵守法规和内部政策至关重要。
数据科学中 ETL 的最佳实践：
为了最大限度地提高数据科学中 ETL 过程的有效性和效率，应考虑以下最佳实践：
The ETL (Extract, Transform, Load) process plays a significant role in data science by facilitating the acquisition, preparation, and integration of data for analysis and modeling. In this article, we will delve into the ETL process specifically within the context of data science, examining its key components and best practices.
Extract:
The first step in the ETL process for data science is data extraction. Data can be sourced from a variety of locations, including databases, APIs, web scraping, sensor data, social media platforms, and more. The extraction phase involves identifying the relevant data sources and retrieving the required data. This may entail querying databases, making API requests, or utilizing web scraping techniques. The extracted data may be structured, semi-structured, or unstructured, and could encompass text, numerical values, images, or other forms of data.
Transform:
The transformation step in the ETL process is critical for data science. It involves cleaning, preprocessing, and manipulating the extracted data to make it suitable for analysis and modeling. This phase encompasses tasks such as data cleaning, missing value imputation, data normalization, feature engineering, dimensionality reduction, and data aggregation. Data scientists may employ various techniques and algorithms during this stage, depending on the nature of the data and the objectives of the analysis.
Load:
The final step in the ETL process for data science is data loading. Once the data has been transformed, it needs to be loaded into a suitable format or structure for further analysis. This can involve storing the data in a database, a data lake, or a specific file format. It is essential to ensure data integrity and security during the loading process, as well as to establish appropriate data governance practices to comply with regulations and internal policies.
Best Practices for ETL in Data Science:
To maximize the effectiveness and efficiency of the ETL process in data science, the following best practices should be considered: