Snow Strikes
in Metis
La Nina winters have returned…
I dunno about you, but with regards to ENSO, I kinda prefer the El Niño years. The cyclones are “over there” in the Pacific and the winters tend to be dry.
Between flu seasons and snow season, if you’re someone considering the Metis Bootcamp, I might suggest you avoid the winter season. Of course, if you ARE such a person, might I suggest you list me as one who recommended you to Metis? Chuckle…
In any case, I was “remote” Wednesday and Thursday this week due to kids down with the flu. Then EVERYONE was remote Friday since the Metis Chicago campus was closed due to the Winter Storm..
Bizarrely, the Pair Programming schedule aligned with this. We have an odd number in the cohort. So as “pairs” go, there is always a leftover. That individual can work solo or they can join any other pair of their choice. I dutifully got in front of my computer Wednesday morning before 9am to identify my pair partner only to find I had none. Then I looked at the assignment for the day…
OK. I have to pause a bit and opine a bit about Pair Programming.
We’re past the initial weirdness where we didn’t know each other and where egos may have come into play. Sure, some are more amenable to working with others. But we’re all in this together. So, my thoughts of the moment don’t have to do with that aspect.
No… it’s the nature of the assignments.
The actual assignments range pretty widely in their difficulty level, their topic, their scope, etc. A lot of the time, they have a “trivial” feel to them. Of course, what’s trivial to one may be novel to another. But you get a sense these are standard teaching exercises or refreshers. And then one comes out of the blue to spank you hard.
For example, this week…
- Monday - SQL exercises (yawn.. refresher)
- Tuesday - standard simple recursion examples. So standard, in fact, that Amy took one look at it and was entirely ready to teach the lesson of memoization because this particular examples was THE “go to” example for that topic.
- Wednesday - Program Batch Gradient Descent from Scratch
- Thursday - Finish Wednesday’s work
- Friday - Run through an entire classification exercise with charts and logistic regression of 6 features.
Do ya see a difference there?!? If not just follow along…
Being “off” Wednesday and Thursday underscored the difference between reading the material and being in a group setting where you can rattle off questions immediately to the instructors. Furthermore there is a clear structure between:
- Reading about certain techniques
- Implementing projects using said techniques
- Running through a complete exercise using said techniques in less than 30 minutes.
Now, I had one HUGE advantage over my peers on Wednesday. They all had to move on once the 30 minutes were up. I had no such restriction. They didn’t fire up Zoom or anything for me. I was just connected via Slack in a loose fashion with the curriculum and such available for me to read at my pace. So the only thing that hindered me from completing the task was my caretaker duties at home.
And… on that note… if I can program Batch Gradient Descent from scratch while enduring the fingernails-on-chalkboard sensation of a teenage daughter whimpering and moaning while fighting off a high fever, I imagine I can endure the pressure of Code Interviews. Mind you, if you’re a prospective employer, might I suggest you peruse my work on HackerRank, GitHub, these projects, etc., and spare me said ordeal. But I’m better prepared nonetheless.
In any case, working with Gradient Descent was interesting because you had to determine things were failing because your code was bad or because your choice of alpha was way off.
I didn’t even realize the plan was to continue Thursday till I again dutifully checked before 9am to see the assignment. Apparently, even then, only a couple pairs had working examples. I’d pat myself on the back except that it’d be an unfair comparison. Sure I got it working… but not in two 30 minute sessions.
Friday was much more fun because the entire class was remote. This made things interesting because of how everyone chose to work together. Several of us had issues with one platform or another. Nate, my pair partner, immediately suggested we stick to nothing more than Slack. This meant no voice. So, actually, we were both just working the problem independently, simultaneously while peppering each other with questions or comments. I was surprised how far I got actually, but there was a strong sense that I was “behind” on the material due to missing lectures. Once the entire class joined via Zoom for review, it became clear our choice was sound. Far too many others wasted too much of their time fighting with getting “connected”.
It was also hilarious to see everyone in their “natural” setting. I must say it seemed several of the students hadn’t really even gotten out of bed yet! We got introduced to a number of pets.
After a couple lectures and a bit of slacking teammates on project work, I had to take the “lunch break” to go out and shovel snow. I just barely finished in time to come back in for a Metis-wide quiz. All four campuses were taking this quiz at roughly the same time. It was declared to be a closed-book quiz. Of course, it’s all on our honor, especially since we (in Chicago) were all remote. But the entire idea is to assess ourselves, both on an individual level and on the levels of each campus and the curriculum/teaching overall. So why bother cheating?!?
We had an hour allocated to the quiz. But I burned through it fairly quickly because for most of it you either knew it or you didn’t. There wasn’t much middle ground. Almost none of it struck me as something I’d never heard of. It wasn’t that kind of “not knowing”. It was more like, oh yeah… that topic… that I’d not yet had the chance to review properly.
There were still a few things where I sort of understood the general idea but hadn’t yet solidified on which term meant what thing. So I did slow down and methodically work through my reasoning so at the very least my answers demonstrated a consistent thinking, even if consistently wrong.
All in all, I’m having a ball, but I’m “at risk” of slipping behind. It’s tough.