These are the lecture and exercise materials related to Introduction to Scientific Machine Learning for Engineers at TUM. Full, readable version at https://tumaer.github.io/SciML
The easiest way to run the exercise notebooks is through Google Colab.
If you want to run them locally, we provide this setup tested on Ubuntu 22.04 and macOS 14:
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
To build the book:
jb build .
To remove all files in the _build
dir:
jb clean .
First, you need to pip install ghp-import
. Then from the master
branch push to GitHub Pages via:
ghp-import -n -p -f _build/html
More info: https://jupyterbook.org/en/stable/publish/gh-pages.html
These are some instructions on how to edit the Jupyter book.
Equations
Name equations uniquely to enable referencing them across the whole book, e.g.
$$y = ax + b$$ (weltgleichung)
Die Weltgleichung {eq}`weltgleichung`.
Do not forget to include an emptry line before and after an equation, unless it is within the text.
Figures
For a proper/referencable figure, use
```{figure} ../imgs/my_figure.png
---
width: 500px
align: center
name: my_figure_name
---
My figure (Source: {cite}`bishop2006`).
```
And then refer to it using
this style: {numref}`my_figure_name`.
Referencing
Cite Bolstad by using
this stype: {cite}`bolstad2009`.
Indentation
If e.g. an equation should be within a bullet point, it should be indented respectively. Otherwise, it breaks the bullet list.
Numbered list:
1. First point: some stuff following equation
$$y = ax + b$$ (weltgleichung2)
2. Some more stuff
Misc
- To put highlighted content, use
> some stuff
on a new line. - To include a simple markdown URL use
[this](https://x.com)
. - To refer to other files use
[Some description](./lecture/x.md)
, or to use their own heading, use[](./lecture/x.md)
. - Inline equations as
$y=f(x)$
. - All functionalities from
.md
apply also to.ipynb
. - Do not prefix images with
ad_
. They are often hidden by add blockers.