Scientific Programming requires the highest performance but we also want to write very high level code to enable rapid prototyping and avoid error prone, low level implementations.
The Julia programming language is designed with exactly those requirements of scientific computing in mind. In this course we will show you how to make use of the tools and advantages that jit-compiled Julia provides over dynamic, high-level languages like Python or lower level languages like C++.
Before joining the course, consider reading the following two blog posts to figure out if Julia is a language in which you want to invest your time.
What will you learn?
First and foremost you will learn how to think julia - meaning how write fast, extensible, reusable, and easy-to-read code using things like optional typing, multiple dispatch, and functional programming concepts. The later part of the course will teach you how to use more advanced concepts like language introspection, metaprogramming, and symbolic computing. Amonst others you will implement your own automatic differentiation (the backbone of modern machine learning) package based on these advanced techniques that can transform intermediate representations of Julia code.
Organization
This course webpage contains all information about the course that you need, including lecture notes, lab instructions, and homeworks. The official format of the course is 2+2 (2h lectures/2h labs per week) for 4 credits.
The official course code is: B0M36SPJ and the timetable for the winter semester 2022 can be found here.
The course will be graded based on points from your homework (max. 20 points) and points from a final project (max. 30 points).
Below is a table that shows which lectures have homeworks (and their points).
| Homework | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Points | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | - | 2 | - | 2 | - |
Hint: The first few homeworks are easier. Use them to fill up your points.
Final project
The final project will be individually agreed on for each student. Ideally you can use this project to solve a problem you have e.g. in your thesis, but don't worry - if you cannot come up with an own project idea, we will suggest one to you. More info and project suggestion can be found here. Check out the list of projects from past to get a feeling for what we expect.
Grading
Your points from the homeworks and the final project are summed and graded by the standard grading scale below.
| Grade | A | B | C | D | E | F |
|---|---|---|---|---|---|---|
| Points | 45-50 | 40-44 | 35-39 | 30-34 | 25-29 | 0-25 |
Teachers
| – | Room | Role | |
|---|---|---|---|
| Tomáš Pevný | pevnak@protonmail.ch | KN:E-406 | Lecturer |
| Vašek Šmídl | smidlva1@fjfi.cvut.cz | KN:E-333 | Lecturer |
| Matěj Zorek | zorekmat@fel.cvut.cz | KN:E-333 | Lab Instructor |
| Niklas Heim | heimnikl@fel.cvut.cz | KN:E-333 | Lab Instructor |
Prerequisites
There are no hard requirements to take the course. We go through the basics of Julia, but we do it rather quickly, so some familiarity with the language is an advantage. If you are looking for an in depth course on the basics of the language, we recommend to check out Julia for Optimization and Learning before enrolling in this course. The Functional Programming course also contains some helpful concepts for this course. And knowledge about computer hardware, namely basics of how CPU works, how it interacts with memory through caches, and basics of multi-threading certainly helps.
References
- Official documentation
- Modern Julia Workflows
- Workflow tips, and what is new in v1.9
- Zero2Hero Julia Workhop
- Think Julia: How to Think Like a Computer Scientist
- From Zero to Julia!
- WikiBooks
- Justin Krumbiel's excellent introduction to the package manager.
- juliadatascience.io contains an excellent introduction to plotting with Makie.
- The art of multiple dispatch
- MIT Course: Julia Computation
- Tim Holy's Advanced Scientific Computing