This is a collection of Jupyter notebooks and tutorials to get you up to speed on lartpc_mlreco3d
. You can view the tutorial page here.
Install Jupyter Book to contribute to the repository (see below how to contribute).
$ pip install -U jupyter-book
Create your Markdown files / Jupyter notebooks. Reference them in the TOC (_toc.yml
).
See https://jupyterbook.org/intro.html for more guidance on how to write your pages.
For better version control, it is preferred that you write your Jupyter notebook using Markdown. A Jupyter notebook written entirely in Markdown needs a YAML frontmatter and looks like this:
---
jupytext:
cell_metadata_filter: -all
formats: md:myst
text_representation:
extension: .md
format_name: myst
format_version: 0.13
jupytext_version: 1.10.3
kernelspec:
display_name: Python 3
language: python
name: python3
execution:
timeout: 240
---
# Your title
```{code-cell}
code that will be executed like a Jupyter notebook cell
```
See https://jupyterbook.org/file-types/myst-notebooks.html for more information.
The tutorials rely on some weight files and small datasets. You can download them all by running the setup.sh
script
from the root of the repository:
$ source setup.sh (optional: path/to/your/folder)
Files are stored by default in the folder lartpc_mlreco3d_tutorials/book/data
.
If you provide a custom path, the script will export that path in the environment variable DATA_DIR
which is used by the tutorials.
Files will be downloaded from SDF. As such, to view the tutorials properly you need access to the SLAC SDF file system. One way to do this is to mount the remote filesystem to your local machine using sshfs:
$ sshfs -o reconnect,ServerAliveInterval=15,ServerAliveCountMax=3 $USERNAME@sdf-login.slac.stanford.edu:$PATH_TO_TUTORIAL $PATH_TO_TUTORIAL_LOCAL
where:
$USERNAME
: your SLACsdf
login ID.$PATH_TO_TUTORIAL
: path to thelartpc_mlreco3d_tutorials
folder inside SDF.$PATH_TO_TUTORIAL_LOCAL
: your desired path to the tutorial folder, in your local machine.
For example, after git cloning this repository in /somewhere/lartpc_mlreco3d_tutorials
, I can make a folder ~/lartpc_mlreco3d_tutorials
in my local Ubuntu, and mount the tutorials folder in SDF to this folder as follows:
sshfs -o reconnect,ServerAliveInterval=15,ServerAliveCountMax=3 [email protected]:/somewhere/lartpc_mlreco3d_tutorials ~/lartpc_mlreco3d_tutorials
This allows one to view files via the file browser of your local machine.
Every time you want to build:
$ jupyter-book build book
You can now open the file _build/html/index.html
to preview your changes.
If you have the right access permissions and the package ghp-import
installed:
$ pip install -U ghp-import
then you can easily update the Github Pages after building:
$ ghp-import -n -p -f book/_build/html
Using the awesome Jupyter Book and Binder. Hosted on Github Pages.