PYTHIA COOK-OFF 2023

PROJECT PITCHES

JUNE 20, 2023

Before we get started

1. Work on what you want
2. These groups can change
3. Ask for help
4. We’ll be editing these slides in real time

Before we get started

• Work on what you want
• These groups can change
• Ask for help

Follow along on Slack!

Radar Cookbook Improvements: Including the Rest of the Ecosystem

​

​

Max Grover (in-person)

Cookbooks as scholarly objects

• Not an individual cookbook, but infrastructure questions (citation and attribution)
• Creation of DOIs best practices, durable links
• Leverage automation to unify and simplify process for contributors

​

Brian Rose (in-person)

3D Visualization

Possible Packages

​

pyvista/pyvista-xarray

​

bjlittle/geovista

​

Borrow From

​

tutorial.pyvista.org

​

But also any geospatial viz!

​

​

​

Bane Sullivan (in-person)

Subsurface Geology

​

​

Thomas Martin (in-person)

• Thin section image analysis using scikit-image & scikit-learn
• Various other ⚒️🪨⚒️ stuff
• Maybe pivot to 3D viz/scikit-learn ML

​

​

The MetPy Cookbook

Convert existing classic gallery of examples to a Cookbook

Drew Camron (in-person)

Interactive ARCO Dataset Analysis and Visualization

​

Use the holoviz ecosystem to develop interactive visualizations of gridded and point-based data, including:

• ERA-5 reanalysis
• Surface (ASOS and mPING) precipitation-type observations
• NCEP High Resolution Rapid Refresh (HRRR) model data

​

Kevin Tyle (in-person) & Alfonso Ladino (in-person)

Climate Variability

Possible packages

​

EOF:

ajdawson/eofs

nicrie/xeofs

zhonghua-zheng/pyEOF

dougiesquire/xeof

​

SVD/MCA:

nicrie/xmca

Yefee/xMCA

numpy.linalg.svd

​

Wavelet:

regeirk/pycwt

jkrasting/xwavelet

​

Other:

rcjwills/lfca

​

Robert Ford (virtual)

VAPOR Python API Cookbook

Nihanth Cherukuru (virtual)

Cookbook for Marine Heatwave Forecast

https://psl.noaa.gov/marine-heatwaves/

Chia-Wei Hsu (in-person)

GPM-DPR Level 2 data analysis

• Explore GPM (global precipitation mission) dataset in Python and 2D visualization
• Interested in preparing GPM products (reflectivity)
• HDF5 python library
• https://gpm.nasa.gov/missions/GPM/DPR

​

​

Jeevan Bhandari (virtual)

1. 2-D plot
2. Cross-sectional plot

Hdf5 library

Cupy-Xarray Examples and Workflows

• Cupy is a drop in replacement for numpy in Xarray data-architecture that allows for leveraging GPUs
• GPU-specific notebook
• Opportunity to explore extending Cookbook infrastructure to support GPU execution

​

​

Negin Sobhani (In-person + virtual)

Additional project ideas & open co-hacking

Packaging project and Repository Best Practices

Infrastructure (& infrastructure accessories)

Room for new pitches!

A cookbook to summarize the outcomes i.e. notebooks of a recent Reproducibility Challenge in Climate and Environmental Sciences (opportunity to test GPU support and multiple software environments) > not sure if this is within the scope of Pythia Cookbooks

​

A cookbook to create and analyze Cloud Regimes (also known as Weather States). Weather states are made by applying k-means clustering to cloud optical depth - cloud top height joint histograms produced by data products such as ISCCP, MODIS and MISR as well as climate models. I also intend to include the option of using a modified k-means algorithm that uses Wasserstein distance instead of euclidean distance.

​

A cookbook to create and analysis Cloud Regimes (also known as Weather States). Weather states are made by applying k-means clustering to cloud optical depth - cloud top height joint histograms produced by data products such as ISCCP, MODIS and MISR as well as climate models. I also intend to include the option of using a modified k-means algorithm that uses Wasserstein distance instead of euclidean distance.