Jim's tutorials for CoDaS-HEP:
- The Scientific Python Ecosystem, Tuesday, July 23, 2019 in 407 Jadwin Hall.
- Columnar Data Analysis, Wednesday, July 24, 2019 (same location).
- Accelerating Python, Wednesday, July 24, 2019 (same location).
How to participate (easy)
Press this button:
The exercises will run on a free public cloud service, which cannot save data and may take a minute or two to start up.
Navigate in the JupyterLab file view (left sidebar) to the desired lesson.
How to participate using CoDaS-HEP resources
Launch a private JupyterLab through SSL as discussed in the CoDaS-HEP setup session (with the
ivukotic/ml_platform_auto:conda image). In JupyterLab, click on "Terminal" and type (first time only):
cd /workspace git clone https://github.com/jpivarski/2019-07-23-codas-hep.git cd 2019-07-23-codas-hep conda env create -f environment.yml
After the first time (e.g. on Wednesday; day 2), do
cd 2019-07-23-codas-hep git pull
Then navigate in the JupyterLab file view (left sidebar) into the
2019-07-23-codas-hep directory to the desired lesson. After you start the notebook, use the Kernel menu, Change Kernel... to ensure that it's in the
To get this software on your laptop
Short answer: on a non-Windows computer, make sure you have conda installed and use the same instructions as the "CoDaS-HEP resources" section above (in your own choice of directory). Then switch to your new environment, manually install JupyterLab, and launch JupyterLab. It will open a web browser window and run on your machine.
conda activate codashep-python-columnar conda install jupyterlab jupyter lab
All of the software installed this way goes into an isolated environment that you can easily delete with
conda remove --name codashep-python-columnar --all
You can also look at environment.yml to pick and choose software to install. Everything in that list can be installed on Windows except ROOT, and everything can be installed using pip instead of conda except ROOT. The packages that must be installed with pip are in the
pip section, and
root_numpy must be installed after ROOT. You can install ROOT manually from here, and be sure to install the latest release, 6.18/00.