Central Iowa Linux Users Group: August 2020 Jupyter Lab

Transcript

1. Jupyter Lab Andrew Denner August 2020 Central Iowa Linux Users Group Video of talk at: https://youtu.be/3_r0LqzMAD8 Source: https://gitlab.com/denner1/cialug-august-2020-jupyter-lab
2. Welcome We will get started shortly
3. Welcome to CIALUG Third Wednesday of the month http://cialug.org Email List IRC/Slack Want to present?
4. Checkin How are you?
6. Linux News
7. Our Presenter • Senior Software Developer • Linux Tinkerer • Twitter: @adenner • Slides will be at: http://denner.co
8. Jupyter Lab Andrew Denner August 2020 Central Iowa Linux Users Group
9. A brief history • Spun off from ipython in 2014 by Fernando Pérez • Jupyter core languages • Julia • Python • R • But now rather agnostic (.net anyone?)
10. Notebooks TNG • Released in 2018 • Same building blocks • Cool new interface!
12. https://mybinder.org/v2/gh/jupyterlab/jupyterlab-demo/try.jupyter.org?urlpath=lab
15. I want more!
16. First Some setup • Follow along: https://gitlab.com/denner1/cialug-august-2020-jupyter-lab • Initial setup on Ubuntu Fossa (20.04) Docker image • docker run -p 8080:8080 ubuntu:20.04 • Or easy button to skip ahead docker run --rm -it –p 8080:8080 registry.gitlab.com/denner1/cialug-august- 2020-jupyter-lab:latest • Please note: this image is not optimized for size or security and is not production ready!
17. Setup cont. • First setup virtualenv and some prereqs • apt-get install git r-base r-cran-devtools nodejs wget python3-pip build-essential libssl- dev libffi-dev python-dev python3-venv • source venv/bin/activate • Verify which env we are in with ‘which pip’ (also note hint in the prompt)
18. Why virtualenv • Not really needed for docker • Think of it as python in a bottle • Other options are conda or just yell YOLO (or docker in our case)
19. Setup cont. • pip install jupyterlab ipywidgets • jupyter lab –ip=0.0.0.0 –port=8080 –allow root • Install R apt-get install r-base r-cran-devtools • In R: • devtools::install_github("IRkernel/IRkernel") • IRkernel::installspec() • GnuPlot • apt install -y gnuplot • pip install gnuplot_kernel • python -m gnuplot_kernel install --user
20. Useful tips • ! Will run in the shell eg. • !pip install packagename • Tab completion is your friend • help() • quickref • Lots of Kernels out there: https://github.com/jupyter/jup yter/wiki/Jupyter-kernels
21. Demo
22. Links • https://plotly.com/python/ipyt hon-notebook-tutorial/ • https://gist.github.com/frfahim /73c0fad6350332cef7a653bc d762f08d • https://jupyter.org/try • https://github.com/jupyter/jup yter/wiki/Jupyter-kernels