Hi again! This week was kind of busy, yet kind of chill. Lets recap.
So what did I do?
I applied to college lol.
But the goals I set in the previous post are kind of unrealistic. I don’t feel bad for being over ambitious. I would much rather reign in my goals then finish a month thinking that I could’ve done more. So here’s 2024, rewritten.
All of the motivations for these goals were explained in the last post.
Continuing on, here’s January, rewritten.
finally, this week!
Most of the goals of the form do x (weekly, daily, monthly) are just going to be melded into my routine as I experiment more. We’ll see.
THANK GOD FOR TEST DRIVEN DEVELOPMENT!
But seriously, I don’t even understand how a project is even possible without TDD. I implemented a couple of new features this week and in the process SOOOO many things broke that I wouldn’t have caught without my tests. Seriously thinking of switching to a new testing framework.
If you don’t know, Test Driven Development is the process of writing tests that validate your code before writing your code. Then write the bare minimum code that works and passes your tests, then refactor and clean up your code!
It’s real strength was shown off after I’d accumulated a library of tests. If I wrote code that broke old code, my tests would catch it, I didn’t even have to stress!
Going on to new features, I added in
Of these four features ESCAPE CHARACTERS were the trickiest to implement. It was hard to do so in a way that didn’t nuke all other functionality, so my code is currently a mess.
| Links are strange because instead of following the * <– open tag | content | –> * format, they’re really two tags. Links look like this [text](hyperlink), this is super annoying and I had to implement a solution I felt less than comfortable with. I’m thinking of getting my code reviewed on reddit to try and improve my solutions to these less obvious cases. |
This is the course I’ll be referring to
So week one was just a more general review of everything I’d learned over these past couple of weeks I’ve spent with ML. Linear algebra, tensor manipulation in python etc. The actual content started of with Linear Classifiers. A pretty basic ML model that can easily be perfectly optimized. In short, it’s just a plane through an arbitrarily high dimensional space. Points are scattered through out the space, and what we learn is a plane that separates classes. A 1-D example would be a classifier trying to figure out if you’re above a certain height by looking at the length of your femur. If your femur lies on a line above the LC, then you’re over the height we trained for, if not, you’re under! They go into a lot more mathematical detail that I’m still trying to understand myself. I’ll let you know how it goes.