Imageomics / Imageomics/LatLonCover

Offer option to download CDL locally for improved performance

Open
#17 0 comments 0 reactions 0 assignees View on GitHub
enhancement
Dominant language
Jupyter Notebook
Stars
4
Forks
0
PR merge metrics
No merged PRs in 30d

Description

Since the CropScape API can be slow to retrieve land-use data, we may consider simply [downloading the most recent CDL](https://www.nass.usda.gov/Research_and_Science/Cropland/Release/index.php) and putting that directly in the local filesystem.

A possibility for implementation could have LatLonCover configured to first look for CDL data in a local directory and then revert to the API. If it is not found locally, there could be a prompt for the option to do a one-time download from the CDL source to improve performance.

Each year is only ~2GB, and this would fit, for example, our use with Andromeda on Hugging Face, as the [hardware](https://huggingface.co/docs/hub/spaces-overview#:~:text=preferred%20hardware%20environment.-,Hardware,-GPU%20Memory) provided in a free space can easily fit this.

Even though Andromeda is not meant to handle a huge number of observations, having better performance retrieving land use data might be needed for larger scale uses of this dependency.

Contributor guide

No contributing guide indexed for this repository

Research direction

Start by reviewing LatLonCover's current CropScape API retrieval path and how local filesystem data could be configured. Done would mean an optional local CDL source, a one-time download from the linked CDL source when data is absent, and fallback to the API.

Written by the indexing model from the issue text.

Assessment

Domain
data, performance
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.