Skip to content
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

Add datasets + samplers tutorial #2455

Open
wants to merge 10 commits into
base: main
Choose a base branch
from

Conversation

hfangcat
Copy link
Contributor

@hfangcat hfangcat commented Dec 6, 2024

Adding a new tutorial for datasets and samplers according to #2418.

Weird points here (to be fixed later):

  1. The ChesapeakeCVPR download link is closed, I used my previous saved zip.
  2. The ChesapeakeCVPR dataset is a bit large to download, which might not be suitable for a beginner tutorial, it would be good if we could provide a subset for the dataset (or use other alternative datasets).

@github-actions github-actions bot added the documentation Improvements or additions to documentation label Dec 6, 2024
@adamjstewart
Copy link
Collaborator

Can you remove the changes to pretrained/transforms from this PR?

@hfangcat hfangcat force-pushed the add-datasets-tutorial branch from 1974bf2 to 25c404f Compare December 7, 2024 15:33
@hfangcat
Copy link
Contributor Author

hfangcat commented Dec 7, 2024

Can you remove the changes to pretrained/transforms from this PR?

done!

@adamjstewart adamjstewart mentioned this pull request Dec 8, 2024
25 tasks
@adamjstewart adamjstewart added this to the 0.6.2 milestone Dec 8, 2024
@adamjstewart
Copy link
Collaborator

This tutorial doesn't really add anything that isn't already covered in the Introduction to TorchGeo tutorial. I would like to go much deeper, including the structure of GeoDataset R-trees, the class hierarchy (including RasterDataset and VectorDataset), how to use Raster/VectorDataset for generic uncurated data, the different splitters available, sampler ROIs, etc.

Also, the tests aren't passing. We can only use toy datasets in our tutorials due to time/space constraints.

@hfangcat
Copy link
Contributor Author

hfangcat commented Dec 8, 2024

This tutorial doesn't really add anything that isn't already covered in the Introduction to TorchGeo tutorial. I would like to go much deeper, including the structure of GeoDataset R-trees, the class hierarchy (including RasterDataset and VectorDataset), how to use Raster/VectorDataset for generic uncurated data, the different splitters available, sampler ROIs, etc.

Also, the tests aren't passing. We can only use toy datasets in our tutorials due to time/space constraints.

I agree with the deeper tutorial, but there is no time to do that for the AGU tutorial now. I fixed the ruff but the dataset (download issue) would be fixed later...

@adamjstewart
Copy link
Collaborator

Well, it can't be merged without passing tests. Also, it doesn't contain any new content that isn't shown in the introduction.

@adamjstewart adamjstewart modified the milestones: 0.6.2, 0.6.3 Dec 8, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
documentation Improvements or additions to documentation
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants