Table of Contents for
Data Wrangling with Python

Version ebook / Retour

Cover image for bash Cookbook, 2nd Edition Data Wrangling with Python by Katharine Jarmul Published by O'Reilly Media, Inc., 2016
  1. Cover
  2. nav
  3. Praise for Data Wrangling with Python
  4. Data Wrangling with Python
  5. Data Wrangling with Python
  6. Preface
  7. 1. Introduction to Python
  8. 2. Python Basics
  9. 3. Data Meant to Be Read by Machines
  10. 4. Working with Excel Files
  11. 5. PDFs and Problem Solving in Python
  12. 6. Acquiring and Storing Data
  13. 7. Data Cleanup: Investigation, Matching, and Formatting
  14. 8. Data Cleanup: Standardizing and Scripting
  15. 9. Data Exploration and Analysis
  16. 10. Presenting Your Data
  17. 11. Web Scraping: Acquiring and Storing Data from the Web
  18. 12. Advanced Web Scraping: Screen Scrapers and Spiders
  19. 13. APIs
  20. 14. Automation and Scaling
  21. 15. Conclusion
  22. A. Comparison of Languages Mentioned
  23. B. Python Resources for Beginners
  24. C. Learning the Command Line
  25. D. Advanced Python Setup
  26. E. Python Gotchas
  27. F. IPython Hints
  28. G. Using Amazon Web Services
  29. Index
  30. About the Authors
  31. Colophon

Appendix A. Comparison of Languages Mentioned

Often when you’re working with a programming language, others will ask why you use that language. Why not use X or Y? X or Y may vary depending on what the person knows and whether they are an avid developer. It is good to understand why they are asking that question, and to think about your reply—why Python? This appendix compares Python to other useful languages so you can answer these questions and gain some insight into our programming choices.

C, C++, and Java Versus Python

When compared with C, C++, and Java, Python is fairly easy to learn, especially for those without a computer science background. As such, many folks who may have started in your same position have built add-ons and helpful tools to make Python more powerful and useful for the data science and data wrangling realms.

As for the technical differences, Python is a high-level language, while C and C++ are low-level languages. Java is high level, but has some low-level qualities. What does this mean? A high-level language abstracts interactions with the computer architecture—that is, it allows you to type code words (say, a for loop or variable definition), which the language then compiles down to code a computer can execute—while a low-level language deals with them directly. Low-level computer languages can run faster than high-level languages and allow for more direct control over a system to optimize things like memory management. High-level languages are easier to learn because most of those lower-level tasks are already managed for you.

For the purposes of the exercises taught in this book, there is no need to manipulate system control or speed things up by several seconds—so we do not need a low-level language. While Java is a high-level language, it has a higher learning curve than Python, and it would take longer for you to ramp up and get started.

R or MATLAB Versus Python

Python has libraries (supplemental code) with many of the same capabilities as R and MATLAB. Those libraries are called pandas and numpy. These libraries handle specific tasks related to big data and statistical analysis. If you would like to learn more about them, you should check out Wes McKinney’s book Python for Data Analysis. If you have a strong background in R or MATLAB, you can still use those tools for data wrangling. If that is the case, Python is a great supplemental tool. However, having all the pieces of your workflow in the same language makes data processing easier and more maintainable. By learning both R (or MATLAB) and Python, you can pick and choose which language you would like to use based on the needs of a particular project, giving you extra adaptability and convenience.

HTML Versus Python

Explaining why you don’t use HTML to wrangle data is like explaining why you don’t put water in a gas tank—you just don’t. It is not made for that. HTML stands for HyperText Markup Language, and is the language that provides the structure for web pages to be displayed in a browser. Just like we talked about in Chapter 3, when we discussed XML, we can use Python to parse HTML, but not the other way around.

JavaScript Versus Python

JavaScript, which should not be confused with Java, is a language that adds interactivity and functionality to a web page. It runs in the browser. Python is divorced from the browser and runs on the computer system. Python has a rich collection of libraries that add functionality relevant to data analysis. JavaScript has extra functionality relating to browser-specific purposes. You can scrape the web and build charts with JavaScript, but not run statistical aggregation.

Node.js Versus Python

Node.js is a web platform, while Python is a language. There are frameworks written in Python similar to Node.js, like Flask and Django, but Node.js is written in the JavaScript language. While JavaScript is predominantly used on the Node.js allows for you to use JavaScript on the backend. If you use something like Flask or Django, you will probably have to learn JavaScript to use for your frontend needs. However, most of the work in this book is aimed at backend processes and larger data processing. Python is more accessible, easier to learn, and has specific data processing libraries already created for data wrangling use. For that reason, we use Python.

Ruby and Ruby on Rails Versus Python

You may have heard of Ruby on Rails, which is a popular web framework based on the Ruby language. There are many for Python—Flask, Django, Bottle, Pyramid, etc.—and Ruby is also often used without a web framework. We are using Python for its fast processing and data wrangling capabilities—not its web abilities. While we do talk about displaying data, if your goal is to build a website, you are reading the wrong book.