nasaharvest / crop-mask

End-to-end workflow for generating high resolution cropland maps
Apache License 2.0
95 stars 28 forks source link

User-friendly interface for intercomparison results #338

Open hannah-rae opened 8 months ago

hannah-rae commented 8 months ago

Context: The intercomparison_script.ipynb notebook creates results for inter-comparing the accuracy of many land cover/cropland maps for a subset of our evaluation datasets (configured in src/compare_covermaps.py). The results are stored in notebooks/intercomparison-results.csv.

Problem: The CSV of results is not very user-friendly to read and understand the comparative performance of maps for a given country/test set. A user-friendly interface could be created to explore the intercomparison results.

Potential Solution: There are many potential solutions that could be brainstormed for this, but potential solutions could include:

Dev-Destructor commented 8 months ago

Hey, how about an SSR website build using Next.js which will read through the CSV and visualize output?

Aniket-Parlikar commented 7 months ago

Interested in contributing to this issue