100 Days of code – 20% complete

I’ve been running into some problems in the last 5 days.

I’ll work on a problem, run into something I don’t understand, hack at it for a while, realize I need to learn more about that topic, and move on to a different challenge. In one way it’s good, because I’m reaching the edge of what I know and what I’m comfortable with. However, in most cases, I’d need to spend a lot of time studying before I can move on. Maybe I’m a stickler for the rules, but I’m pretty sure it’s “100 days of code” and not “20 days of code, 80 days reading a statistics book in front of a keyboard.”

I found that I’m running into problems in these areas:

  1. Statistics and other math concepts
  2. “Advanced” Python concepts, like list comprehensions, classes and generators
  3.  Lack of a long-term project

Here’s how I’m addressing these problems:

Math knowledge

I have attended a couple data science/Python meetups from Thinkful Raleigh. Those presentations pretty much revolved around using a Jupyter notebook to solve a particular problem, and using that problem to further discuss data science as a whole. I asked whether there’s any good intro to statistics books they could recommend at the first meetup, and they pointed me to An Introduction to Statistical Learning. That book’s pretty good so far, but it’s definitely not a page-turner. I’m also watching videos from Brandon Foltz on YouTube about linear regression.

Advanced Python concepts

If you can do everything with normal functions, why use classes and generators, right? I’ve been playing around with these concepts for a while, but it’s always been more of a copy/paste thing than a real understanding of how these features work. I think my best bet here is to go through some projects or challenges that I’ve already completed and write another solution that includes one of these features.

Long term project

This is probably where I’m struggling the most, because I don’t have any ideas right now. I think part of it is that I don’t think of automation as the best tool for very many problems right now. That’s probably because I still see programming even simple scripts as taking a decent amount of effort. I hope that once I’m better at coding, I’ll come up with some ideas. I’ll keep looking for other projects for now, though.

So, the plan for the next 5 days is to go through some previously completed challenges and force a list comprehension, class, or generator into each solution. While that’s going on, I’m studying statistics in my off time so I can move forward with other problems. And while everything else is going on, I’m looking out for projects.

I’ll update this again in another 5 days. Happy Thanksgiving!