Hello! Welcome to my jupyter lab template. This template it has been created in order to have a easy and fast way of create some notebooks locally and work with them.
This repo is always evolving so, if you find some problems or you want to have some extra features feel free to open a new issue or start a discussion.
To use this proyect you only need to have installed Docker
and use a Unix base
system like: ubuntu, MacOS, wsl.
There is a few forms to run this proyect that are explained bellow, but first you need to know how the directories are structured.
data/
: here you can put your files to work with on the notebooks. All the files put in there will be ignore by git.notebooks/
: here you can put your.ipynb
files to work with. It's important to have it the without directories to be able to usebuild
andrun-and-build
scripts.
If you want to start up a Jupyterlab server, connect into it to start and start working with notebooks just run:
sh run.sh
This scrip just create a .env
file, a Docker image and run it with some parameters.
:warning: There is a problem right there: there is not ssl certificates for the local
server and it shows a message saying that is not secure, just ignore it for now. I will
try to fix it
If you want to export your notebooks in notebooks/
directory as the are to html
format without run it you can use:
sh build.sh
This script creates on the same way as the run.sh
but on detach mode, run a command
into the docker container and stop it.
If you want to run your notebooks under the notebooks/
directory and export the
output into html
format you can use:
sh run-and-build.sh
This script creates on the same way as the run.sh
but on detach mode, run a command
into the docker container and stop it.
I know that it needs to be improved to be able to work on https connections and avoid the initial messages on startup. I will fix it in some point. Also. I know that it could be a good idea to have a makefile instead of multiple scrips and migrate the requirements file to manage dependencies to pipfile. If you know how to do it, open a pull request 😉
-
👨💻 All of my projects are available at criskrus.com
-
📫 How to reach me [email protected]