A.3. IDE
Now that you have Python 3 and NLPIA on your machine, you only need a good text editor to round out your integrated development environment (IDE). Rather than installing a complete system like PyCharm by JetBrains, we prefer individual tools with small developer teams (team of one for Sublime Text) that do one thing well.
Tip
Built by developers for developers is a real thing, especially if the developer team is a team of one. Individual developers often build better tools than corporations because individuals are more open to incorporating code and suggestions by their users. An individual developer that builds a tool because they need it usually builds a tool that’s optimized for their workflow. And their workflow is pretty awesome if they build tools that are reliable, powerful, and popular enough to compete. Large open source projects like jupyter are awesome, too, but in a different way. They’re usually extremely versatile and full-featured, as long as they don’t have a commercially licensed fork of the open source project.
Fortunately the tools you need for your Python IDE are all free, extensible, and continuously maintained. Most are even open source, so you can make them your own:
- Sublime Text 3 (www.sublimetext.com/3) text editor with Package Control (https://packagecontrol.io/installation#st3) and Anaconda (https://packagecontrol.io/packages/Anaconda) linter plus auto-corrector
- Meld merge tool for Mac (https://yousseb.github.io/meld) or other OSes (http://meldmerge.org)
- ipython (jupyter console) for your Read → Eval → Print → Loop (development workflow)
- jupyter notebook for creating reports, tutorials, and blog posts, or for sharing your results with your boss
Tip
Some phenomenally productive developers use a REPL workflow for Python.[1] The ipython, jupyter console, and jupyter notebook REPL consoles are particularly powerful, with their help, ?, ??, and % magic commands, plus automatic tab-completion of attributes, methods, arguments, file paths, and even dict keys. Before Googling or overflowing your stack, explore the docstrings and source code of the Python packages you’ve imported by trying commands like >>> sklearn.linear_model.BayesianRidge??. Python’s REPLs even allow you to execute shell commands (try >>> !git pull or >>> !find . -name nlpia) to keep your fingers on the keyboard, minimizing context switching and maximizing productivity.
That’s you, Steven “Digital Nomad” Skoczen and Aleck “The Dude” Landgraf.