microsoft / torchgeo

TorchGeo: datasets, samplers, transforms, and pre-trained models for geospatial data
https://www.osgeo.org/projects/torchgeo/
MIT License
2.79k stars 353 forks source link

Esri2020 dataset download HTTP Error 409 #2385

Open giswqs opened 4 weeks ago

giswqs commented 4 weeks ago

Description

The Esri2020 dataset is not downloadable.

HTTPError: HTTP Error 409: Public access is not permitted on this storage account.

Steps to reproduce

from torchgeo.datasets import Esri2020
root = os.path.join("datasets", 'Esri2020')
dataset = Esri2020(paths=root, download=True)

Version

0.6.1

calebrob6 commented 4 weeks ago

Thanks @giswqs! This looks like a case where the Planetary Computer didn't require .sign(...) for some datasets, but now does for all. However, it looks like this dataset is deprecated and will be removed next month -- https://planetarycomputer.microsoft.com/dataset/io-lulc.

calebrob6 commented 4 weeks ago

We should change this dataset to pull from https://planetarycomputer.microsoft.com/dataset/io-lulc-annual-v02 (and maybe rename it as this isn't an "Esri" dataset)

adamjstewart commented 4 weeks ago

What's the license? Can we redistribute it on HF so we don't need a signature and can guarantee future availability?