-
Notifications
You must be signed in to change notification settings - Fork 378
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Rename VisionDataset to NonGeoDataset #627
Conversation
I don't mind the rename, although the set of things that are "NonGeoDatasets" is quite large :). I'd like to merge this last before 0.3 so that existing PRs don't have to rebase.
If it is easy to do, maybe. If it is difficult at all, no. |
Super easy to do, like 1 line of code to make it work and maybe 4 lines of code to add a deprecation warning. |
Cool, then why not! |
7e11a67
to
afde40c
Compare
Think I finally got this working. The only files you need to review are |
80e08d4
to
25c9387
Compare
* Rename VisionDataset to NonGeoDataset * Keep VisionDataset but add DeprecationWarning * mypy fix * More fixes * More fixes * cast types * Undo cast * Fix usage in test * No idea why... * Update more datasets
This PR renames the following classes:
VisionDataset
->NonGeoDataset
VisionClassificationDataset
->NonGeoClassificationDataset
TorchGeo was never intended to be a computer vision library, it is for all geospatial data types. Originally, we mostly had computer vision (satellite/aerial/drone imagery) datasets, but we now have multiple non-vision datasets:
Question: should I keep
VisionDataset
andVisionClassificationDataset
aliases with deprecation warnings to help ease the transition? I'm not sure how widely used we are or how stable our users expect us to be. We're still in alpha release, but users don't necessarily realize the implications of that.