First week at Metis

My first week at Metis has just breezed by. While I am not new to data analytics and statistics, I am new to data science and Python. For me, this past week has been a mix of a few things old and familiar, and a lot of things new and surprising. We’ve covered lots of introductory grounds on Git and GitHub, pair programming, Pandas, complexity, Matplotlib, command line basics, Seaborn, and, of course, many more Python topics. Having barely coded 10 lines of Python code before this boot camp, I am now a big fan of Python. It seems to have the capacity to do anything, from data science to web development, and when used with Anaconda and iPython notebooks, it can even replace spreadsheet applications for business analysts.

What is Metis?

Metis is an accredited immersive data science bootcamp experience that combines traditional in-class instruction in theory and technique with real-world data project work.

Highlights of the week

Git and GitHub

Having used git on and off for a few years now, I am fairly comfortable with the basic git commands. But throughout the years I have only ever used one GitHub workflow, which consisted of pushing and merging a feature branch onto a single remote repository on GitHub. The preferred GitHub workflow at Metis is to have two remotes, where one is the original upstream repository, and the other is my fork of that repository. The first GitHub workflow pattern has fewer steps and is a bit easier, but the latter offers greater flexibility and is a bit safer.

Pair programming

Everyone whom I’ve asked in my cohort loves pair programming! Almost every morning over the past week, we’ve paired up to tackle an interesting coding puzzle, code it up and figure out its complexity.

Pandas, Matplotlib and Seaborn

Pandas is just amazing. It makes importing data from various formats so easy and fast, and has so many built-in features. Matplotlib is handy for plotting graphs inline in Python notebooks, and Seaborn makes beautiful graphs out of the box.

Running iPython notebooks on a PC

I’ve decided to try to use a PC throughout the boot camp. As a precaution, I’ve also set up an Ubuntu virtual machine, in case I hit a wall with Windows and have to switch to a *nix OS. I’ve read a lot of warnings against running Python on Windows, but decided that many of them are a bit too nebulous and/or outdated to take literally. Those warnings may well turn out to be right, but in this case I would rather establish that empirically than go on faith. From what I’ve seen so far, Anaconda for Windows, along with Anaconda Navigator, take care of many of the frequent obstacles people encounter when running Python on Windows. I might run into problems with some Python libraries that may not be very Windows-friendly, but I’ll try to cross that bridge when I get there. So far, using Python on PC with Anaconda has been smooth sailings (although it does sometimes require being comfortable with Windows terminal).