Skip to content
New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

Support for Multi-band rasters #58

Open
ozak opened this issue May 28, 2020 · 10 comments
Open

Support for Multi-band rasters #58

ozak opened this issue May 28, 2020 · 10 comments

Comments

@ozak
Copy link
Owner

ozak commented May 28, 2020

Seems it would be useful to handle multi-band rasters. May need first to move to rasterio and then a rethink on how to handle and change functions accordingly. Should resolve issues #26, #39, #54.

@snowman2
Copy link

snowman2 commented Oct 5, 2020

I wonder if rioxarray might help with the transition?

https://github.com/corteva/rioxarray

@ozak
Copy link
Owner Author

ozak commented Oct 7, 2020

Looks like an interesting project. Seems there is some overlap between what this package does and that one.

@snowman2
Copy link

snowman2 commented Oct 8, 2020

They do attempt to solve similar problems. I was thinking it might simplify the transition to use rasterio. Also, it uses Xarray, which I also thought might be useful for solving the multi-band/multi-subdataset support.

@pessimo
Copy link

pessimo commented Sep 7, 2021

just found that georasters.from_file gave the wrong meta info when handling multi-band raster file, maybe currently we should raise exception?

in my case, read a 3 bands(rgb) raster file gave wrong xmin, xmax, so shape of raster file was (3, 256) instead of (3, 256, 256)

@ozak
Copy link
Owner Author

ozak commented Sep 7, 2021

Indeed that is a major issue right now. No multilayer support. I do not think it should be difficult to add a flag to select band for importing once the move to rasterio has been done. I just have not had the time. Does anyone want to try to tackle this? Help would be greatly appreciated.

@volkanyazici
Copy link

In my study, I work with 12 bands. When I tried to read tif file as pandas dataframe with the code df = data.to_pandas() I get an error like this ValueError: Must pass 2-d input. shape=(12, 1257, 1065). And then I found this issue. So, can you tell if there is a multi-band support yet? Thanks.

@ozak
Copy link
Owner Author

ozak commented Mar 22, 2022

Sorry not yet. I haven't had the time to do this. It may need quite a bit of rewriting of the whole package to ensure things are being computed the right way. It's one of the issues where help would be more than welcome.

@volkanyazici
Copy link

I would like to point out that since I am new to python, I am not qualified to edit the codes. I hope someone will spare their help to improve this beautiful package.

@shelleyg-bit
Copy link

I would love to work on this feature. I have been looking to contribute to open source, this seems just the right amount of challenge. I came here because I wanted to analyze a multi-band raster using pandas.

@ozak
Copy link
Owner Author

ozak commented Apr 23, 2022

Thanks @shelleyg-bit That would be great! I think the way to tackle this will need us to first move the import to rasterio which will allow importing one or more bands. Then I think it's just deciding whether to create a new MB ie multi-band class or change the code to ensure that things works fine. My own preference would be to create a new MB class. Do you want to start a new branch for the rasterio changes first. Then we can try this part?

# for free to join this conversation on GitHub. Already have an account? # to comment
Projects
None yet
Development

No branches or pull requests

5 participants