"Writing Code in Econ" Series
You'll need accounts with the following websites if you don't already have one. These are accounts that you'll likely have for a long time, so don't worry about tieing them to your work email.
You'll also need a bunch of new software.
Many built-in text editors that come with social science software are okay. Many are hot garbage. If you're going to spend any significant amount of time writing code, whether it's Stata, R, Matlab, Python, or Latex, you will be productive using a single, professional-grade text editor. This is true for two reasons.
First, real text editors have a number of features designed to make your life easier, like auto-completion, Git integration, syntax checking/linting, code folding, project-level search and replace, etc. Some built-in editors have some of these features. Most have few or none.
Second, using a single editor across all languages means that as you get better in one language, you get better in all of them. You have one set of keyboard shortcuts to learn, one set of options to set, one set of tricks to master.
If you're not already attached to a serious editor (Vim, Emacs, Sublime Text), I recommend you start with Visual Studio Code.
Download Git here. If you're not on Windows you might have other ways of getting Git on your computer.
You will be asked if you want Git to be in your PATH and if you want Git-related tools in your PATH. Choose to have Git in your PATH but not the tools.
More info on setting up and using Git can be found here.
This is only for people on Windows because Windows has awful command line interfaces (consoles). CMDer is an attempt to fix that. Download it here.
Also, once you have Git installed, you can use Git bash through CMDer to have
access to all the amazing tools there like
Python is a very powerful programming language that we will use for statistical analysis and cleaning data. There are a number of extensions to Python, called packages, that help with this. Keeping track of all these packages can get difficult (especially on Windows). Anaconda solves this problem by keeping track of packages automatically.
Download it here.
You can install regular Anaconda, which will automatically install a lot of
packages, or you can install "Miniconda" which will not install any packages
until you ask it to. In either case, I recommend you install it to
C:\Anaconda if you are on Windows.
There are lots of articles online about getting started with Anaconda. This is a random one I found that looks okay.
These are optional or as the need arises.
- Sourcetree. This is an easy way to visualize Git repositories.
- Trello desktop app. As long as you're not on Windows 7, you should be able to get the Trello desktop app, which I think is alittle nicer to use regularly than the in-browser version.
- TeXLive. This is basically Anaconda for LaTeX. Set install size to the biggest available; set default paper to letter; add latex to your path; do not install the offered editor. Be aware that this takes 30+ minutes to install, depending on your internet connection.
Next in "Writing Code in Econ": Project Organization