Courses I've Taken

A compilation of the courses I've taken at CMU

I've seen a number of CMU students write about their courses, so I'd thought I'd do the same. For some context, I came to CMU with the intention of majoring in Physics, which explains the many science classes that I initially took as well as my lack of programming and formal math experience.

! : Courses that I particularly enjoyed

? : Courses that I found very difficult

Fall 2024

• 10-315 Introduction to Machine Learning (SCS Majors)

  This course had pretty terrible course infrastructure. Take 10-301 instead if you want to learn machine learning.

• 15-312 Foundations of Programming Languages

  I think it's unfair for me to write a review for this course, given my previous experience with other PL courses and having TAd 15150 for five semesters (the two courses share a lot of similar ideas). I more or less breezed through the content I was familiar with while really struggling with the material that was more foreign. I think the main benefit of this course is to provide rigorous definitions which motivate how and why programming languages are built in specific ways (lazy vs eager evaluation, dynamically vs statically typed, etc.). It's something I've thought about before, but actually being able to factually motivate language design was pretty interesting. I wasn't a fan of the proofs in this course—it was never clear how much I was allowed to assume, which I thought was important given how much of the course relies on precise inductive rules. I'd still recommend this course, as I think it's a really solid introduction to programming language theory.

• 15-441 Networking and the Internet

  This course also had terrible course infrastructure. This seemed more like a rare occurrence, though—it'll hopefully be better in future semesters.

• 80-100 Introduction to Philosophy

  I liked this course more than I thought I would. To be honest, I came into this course expecting that much of the material would be "philosophical nitpicks" that I wouldn't really care for. But some of the readings I actually found interesting and definitely impacted how I think about the world. I found that the most valuable thing from this course (and other philosophy courses, probably) was learning to analyze arguments that aren't necessarily "correct". Being able to break down an argument into its fundamental premises, then analyzing each one and their logical connections is a technique that seems pretty useful to have. There being two readings a week, though, the workload isn't trivial. It would've probably been better to take this course as a freshman.

Reflections:

This was my busiest semester so far, partially due to grappling with pretty awful course infrastructure in multiple courses (I don't think it's appropriate to discuss what sorts of infrastructure I mean here). This was also the semester where I decided that I might want to consider pursuing compilers study in the future, and that my plans have shifted from taking systems courses to programming language theory courses.

Spring 2024

• 15-411 Compiler Design (!)

  There's a lot to say about compilers, given the amount of work that goes into it. The average workload was about 18 hours a week, with the workload peaking at about 40 hours a week. But to justify why this isn't a "very difficult course," I don't think that there's anything unintuitive about compilers. Every phase of the compiler has a clear purpose, and most phases follow specific rules defined inherently by the programming language. The time commitment mainly comes from the implementation and debugging.

  I think the main theme to take away from compilers is "incremental software development". The first four labs have you implementing different features of C0 (a safe subset of C), like conditional expressions or memory. Each addition changes every phase of the compiler, which occasionally involves lots of refactoring and rewriting. This indirectly emphasizes questions about future-proofing code (which as it turns out, usually isn't worth the effort) and good coding practices (like avoiding wildcards on data types likely to change over time). Something I particularly liked was the design aspect—there (intentionally) isn't a lot of guidance on how to build your compiler, so there's a lot of thought put into how data types should look and how data should flow from one phase of the compiler to another. This is an experience that feels unique to any other CS courses I've taken, though maybe software engineering courses emphasize design topics like these. But it was a valuable experience to have so much control over the codebase.

  Unlike most of my peers, I disliked L5, the lab where you apply optimizations to compete against gcc's execution speed. Implementing each optimization involved digging through some research paper to find the correct algorithm, then debugging for an indefinite time because optimizations change your program in opaque ways. It was also really stressful how the autograder was broken for some time, meaning that it wasn't ever clear as to when we had achieved a good enough score. Nevertheless, I was proud of the work I was able to accomplish for L5.

  What I liked most about the course, though, was how I could finally take a lot of ideas from functional programming and apply them in a context where it finally felt like it mattered. I think this was due to the need to write correct code and having complex transformations on recursive data structures—both of which a functional way of programming provides in a really clean way.

  Overall, I'd highly recommend this course, given that you have a reliable partner to work with, and have the time to invest. Also, I'd only recommend Rust if you have prior working knowledge of the language (I tried this and greatly struggled as a result).

• 33-456 Advanced Computational Physics (!)

  This course has less to do with physics and is moreso a course on how to cope with and effectively use large amounts of data. It's a broad survey on techniques in parallelism, big data, and machine learning. Some homework assignments have a very "open-ended" nature, which allow for a lot of discovery when examining different data sets. The course is also pretty easy—most assignments only took about 30 lines of code and took no more than a couple hours. I'd highly recommend "advanced" computational "physics" to anyone who wants some exposure to big data and its applications in the real world, or alternatively, anyone looking for an easy science elective.

• 36-226 Introduction to Statistical Inference

  All I got out of statistics is that my algebra autopilot needs work, and that I'm terrible at memorizing definitions.

• 98-008 Student Taught Courses (StuCo): Intro to Rust Lang

Reflections:

I took on a more relaxed course load, mainly due to needing to put more time into work for CS Academy. This probably ended up being the right call, as compilers took more time than I expected. I was also able to spend the extra time preparing for interviews for internships. I've also realized that I've taken all the courses that I've wanted to, so in future semesters, I'll likely take courses in different fields of computer science (breadth over depth) in order to discover what else I might be interested in.

Fall 2023

• 15-317 Constructive Logic

  I strongly considered putting this as a course I enjoyed, but ultimately decided against it. I think it's really cool to see how these carefully designed inference rules allow the construction of derivation trees that end up working out beautifully. But also, actually writing these derivation trees out was a major pain. Topics like Curry-Howard isomorphism made the connections to logic systems to concepts I've previously taken for granted feel really rewarding. On the other hand, it felt like I was missing out on a lot more due to some gaps in knowledge (I still don't know what a monad is). The course is really easy, relative to many of the other CS courses at CMU. But maybe it's too easy—I feel it's designed for people to take solely to satisfy their Logic and Language elective. So I'm not entirely sure how to feel about this course, but I think I'll appreciate it more once I've taken other PL courses. Also, I never want to touch Prolog again.

• 15-451 Algorithm Design and Analysis (?)

  This course combines skills from previous CS core courses in a way that makes it so that you're able to both come up with algorithms and also rigorously prove said algorithms work. I think my main gripe is just how much content is covered throughout the course. I always felt as though I was behind, and with homeworks being so difficult, it was hard to justify thoroughly learning the course content each week when I should instead be trying to solve the problems. Exams were pretty brutal, which is probably true for every CS core course anyway. I enjoyed the programming assignments, as it's fun to actually implement the algorithms to see that they actually work. The course content is probably not very relevant for me (hopefully I never see splay trees during an interview) but the algorithmic thinking that's taught is always nice.

• 15-445 Database Systems (?)

  I really didn't like this course. I think it's a useful course in that I was exposed to a wide variety of algorithms and data structures (since database systems are all about optimizations). But I didn't inherently care for the content, so it was hard to actually remember what was taught during lectures. The worst part of the course are the projects. The project write-ups were fundamentally unclear and were sometimes just incorrect. The codebase was an absolute nightmare to read (which I suppose is a useful skill, given the amount of "bad code" out there). It was also awful that projects depended on previous ones and that their test cases were very non-exhaustive, which meant a bug in project 4 could stem from project 1. I also despised the exams—they made it feel like I needed to memorize everything, and the fact that every question was multiple choice meant that you can't ever demonstrate any sort of understanding. Overall, this course was a miserable experience, but I'd still take it again mostly because of how useful C++ and SQL are to know.

• 36-225 Introduction to Probability Theory

  A pretty easy course for learning the basics of probability. I'd already been exposed to half the material due to CS courses like 15251 and 15210, and the other half was a lot of plug and chug, so it's hard to say how valuable this course actually was.

• 84-405 The Future of Warfare

  With the only assignments being an essay, four take-home quizzes, and a group presentation, I didn't really get anything out of this course.

Reflections:

This was too much work. I decided that I would try a semester where I knew I'd be spread thin timewise in order to see if I could handle it. Turns out, I couldn't—but I also think a major reason for that was because I didn't care for many of my courses. I've determined that in future semesters, I'd only take courses that I'm genuinely interested in taking.

Summer 2023

• 15-213 Introduction to Computer Systems

  I have mixed feelings about this course. On one hand, it's absolutely made me a better programmer. There's a lot of focus on writing code that's readable and modular, which helped a lot when it came to debugging. TA code reviews were also nice in helping point out places where I could improve. On the other hand, I came out of this course feeling as though I still had a shallow understanding of many of the topics in the course. Given that it's an introductory course, this makes sense, but I do wish there were more focus on the content and less on the programming assignments.

• 70-311 Organizational Behavior

  A pretty good course for learning how individuals and organizations "behave" in different settings. I was a fan of the social psychology that was covered in the later portions of this course, so I'll probably take a psychology course later on.

Reflections:

Pretty relaxed overall. My courses didn't take up too much time, which gave me a lot more time to focus on TAing and CS Academy. During this time, I also learned some basic web development using The Odin Project (which was also when I made this website). This is also the semester when I transferred into SCS 🥳.

Spring 2023

• 15-210 Parallel and Sequential Data Structures and Algorithms (?)

  This was the course meant to be the standard "algorithms course", with a lot of focus on parallelism. Homeworks and exams were brutal, but the course material was pretty interesting, as it built a lot on the material from previous CS courses. Other than its lack of structure, it was a pretty good course in teaching algorithmic thinking (which was surprisingly useful for leetcode).

• 15-251 Great Ideas in Theoretical Computer Science (?)

  This course ended up eating up a large portion of my weeks (did weekly homeworks Sunday night to Wednesday night). Homework was brutal, but the course felt structured in a way that worked for the students to succeed. I felt like this course was the most I had learned from thus far, which was nice to see given how much time it took me.

• 33-232 Mathematical Methods of Physics

  A course meant to prepare physics students for the math in junior-level physics courses. Luckily, I had essentially covered most of the math that I needed from 21-241 and 21-259. As it turns out though, taking a surface integral will never be not challenging.

• 76-270 Writing for the Professions

  A pretty typical writing course. It's known as the "resume class", but there's also a lot of focus on writing in a manner that allows readers to "satisfice", which was pretty useful to learn about.

• 38-230 ENGAGE in Wellness: Looking Inward

  To demonstrate my improvement in overall wellness, I wrote a 5-page essay talking about my crippling addiction to Wordle.

• 33-202 Physics Sophomore Colloquium II

Reflections:

Overall, a semester with a lot of work, but came out of it feeling somewhat accomplished. Courses, as well as TAing and CS Academy took up a lot of time, but I think I was able to manage my time better relative to previous semesters. Plus, I enjoyed the actual content in my courses, which gave me a lot more motivation.

Fall 2022

• 33-211 Physics III: Modern Essentials (?)

  Painful. Homework was due four times a week, which made it feel like I was working on this class all the time. Exams were also pretty brutal. As it turns out, reasoning about special relativity and basic quantum theory is unintuitive. I wasn't quite sure how much I actually got out of this course, and it also made me realize that I definitely wasn't looking forward to taking Quantum Mechanics, or any junior-level physics class for that matter.

• 33-231 Physical Analysis

  Differential equations for the Physics students. I really enjoyed the way that it was taught. Due to the fact that it was for physics majors, there was a much larger focus on how differential equations applied to physics and less about the rigorous proofs behind why the math worked. I initially wanted to do physics due to the pretty relationship between calculus and introductory mechanics, so learning another tool to expand my vocabulary of ways to describe physical systems was cool.

• 33-241 Introduction to Computational Physics (!)

  This course was effectively "applied numerical methods" with physics being the main motivation. We were able to model some very interesting systems in this course using numpy and matplotlib (Python packages). My only qualm was that I wish I had taken this course as a junior rather than a sophomore. Some of the math (Fourier transforms and probability distributions) and physics (quantum mechanics) I hadn't seen yet, which made the course significantly harder than it probably should have been.

• 09-105 Introduction to Modern Chemistry I

  A pretty easy course. I mainly relied on my knowledge from AP Chemistry to get through it.

• 82-137 Chinese Calligraphy: Culture and Skills

  A super laid-back course. I probably would've enjoyed this course way more in a less strenuous semester (as it turns out, thinking about physics while trying to write Chinese characters wasn't very relaxing).

• 99-250 Seminar for Peer Mentors

• 33-201 Physics Sophomore Colloquium I

Reflections:

This course load was brutal for a number of reasons. Trying to balance my courses with work with CMU CS Academy and as an SI Leader was difficult. More importantly, I didn't enjoy many of the courses I was taking this semester. This semester was sort of the "last chance" for physics, and I was pretty certain that I was about ready to try and switch majors.

Summer 2022

• 15-122 Principles of Imperative Computation (!)

  This course does a good job of introducing the data structures that I'd need to know, as well as presenting the idea of reasoning about the correctness and safety of imperative programs. I liked this course a lot since it unveiled a lot of the "mysteries" from 15-112 (an example being how aliasing in Python lists ended up being pointers in memory). I also got significantly better at debugging in this course due to having to reason about memory in C. The course was a lot of work during the summer (30-40 hours a week), but I don't regret taking it. Overall, I felt like a way better programmer as a result.

• 15-150 Principles of Functional Programming (!?)

  This was (and will likely continue to be) my favorite course at CMU. Being able to write recursive code and prove its correctness via inductive reasoning was a tricky concept to initially understand, but getting past that initial barrier made the experience feel extremely worth it. 15-150 highlighted the beauty of abstraction in code, which presented a unique way of thinking. Functional programming showed me a "beauty" in computer science, similar to how other disciplines such as physics or math can also be “beautiful” at times. λ Functions are values λ !

Reflections:

Honestly, I had a lot of fun this summer. I wasn't so bogged down by other commitments, so I was mainly able to just focus on classes and work for CMU CS Academy. Taking 15-122 and 15-150 simultaneously was also an interesting experience—it highlighted two drastically different ways to reason about programs. I ended up feeling a bit sick of programming by the end of the summer, and was sort of looking forward to taking some courses that were less coding intensive.

Spring 2022

• 15-112 Fundamentals of Programming and Computer Science (!)

  Probably the most well-known course at CMU, 15-112 definitely lives up to its name. I went into the course having no knowledge of how to program and came out with a full-fledged project in Python (though I'm not particularly proud of it…maybe I can go back and add more to it at some point). The computational thinking that I learned was incredibly valuable for later courses that I would take.

• 21-127 Concepts of Mathematics (?)

  The typical "introduction to discrete mathematics + other topics" course. 21-127 ended up being really hard for me. I hadn't written a mathematical proof before taking this course, so I struggled figuring out how to formalize my ideas. Also, finding the idea for some of the homework problems was brutal. Looking back, however, this was probably due to me just staring at most problems, rather than writing out different ideas and seeing where they led. This course gave me some basic knowledge in formalizing mathematics, which ended up being useful for a lot of other courses.

• 21-241 Matrices and Linear Transformations (?)

  This course was brutal. For some reason, the proofs that you were expected to write in this course ended up being more difficult than the ones with 21-127 (the class that's meant to teach you how to write proofs). Most of the material went way over the head, but by the end, I had a high-level understanding of the basic concepts of linear algebra.

• 33-152 Matter and Interactions II (?)

  Physics 2 for Physics majors. This course ended up being challenging, due to its assumption in being fluent in vector calculus and also mixing in a bit of linear algebra and probability. I came out of this course realizing that I disliked E&M and was probably in the wrong major.

• 33-104 Experimental Physics

  A laid-back and a bit of a tedious course. I learned some basic Excel to optimize doing calculations, which was probably useful.

• 57-299 Bagpipe and Drum Band

  Fun! It was nice having a "required" commitment that wasn't purely academic. Practices being at Monday 7pm - 9pm was tiring though.

  Reflections:

  This semester was intense for me. I greatly underestimated how much time some of these courses would take (especially the math courses). It was around this time that I realized that I probably wasn't going to be too happy as a Physics major, so I started looking into other majors, like CS and ECE.

Fall 2021

• 21-259 Calculus in Three Dimensions

  A useful course to learn the basics of vector calculus. Looking back, I do wish that this course was a little more rigorous, as most theorems were stated without being proved. As it turned out, my retention rate for this course ended up being quite poor.

• 33-151 Matter and Interactions 1

  Physics 1 for Physics majors. A bit boring, given that I took introductory physics in high school. This course had a final project with a computational portion using Jupyter notebook, which I had a lot of fun working on.

• 76-101 Interpretation and Argument

  The typical "first-year writing" course that all CMU students need to take. This course ended up being surprisingly easy. I somewhat wish that I received more feedback, as I didn't come out of course feeling as though my writing skills had improved.

• 03-121 Modern Biology

  Didn't get anything out of this course for a number of reasons.

• 21-295 Putnam Seminar

  A laid-back course to help students prepare for Putnam. This course ended up being pretty useless for me, as I didn't have any mathematical competition experience prior to taking it.

• 38-101 EUREKA!: Discovery and Its Impact

  The "introduction to college class" for the science students at CMU.

Reflections:

Overall, this ended up being a pretty relaxed semester. I had the opportunity to make friends and get involved with a few extracurriculars, which was pretty nice. Looking back, it might've been nice to drop Putnam Seminar and replace it with a more involved course.