How to match .h5 images and CSV

by Einstellung - opened Apr 20, 2023

Apr 20, 2023

•

edited Apr 20, 2023

Hello @rvignav , we would like to use this dataset in the jax diffusers sprint.
We tried to load with datasets.load_dataset() but got the following error "Could not convert {table.schema} to {features} because the column names do not match."
So we cloned the repository and read the .h5 file, now we'd like to know how to get the ids for each image and match them to the metadata.
any clue will be useful
Thank you

rvignav

Rajpurkar Lab / Harvard Medical AI org Apr 20, 2023

Hi, I'd recommend downloading the dataset from PhysioNet at this link. There shouldn't be any errors there. However, feel free to re-open the thread if that doesn't work!

rvignav changed discussion status to closed Apr 20, 2023

Einstellung

Apr 20, 2023

I mean we've already download and read the .h5 and the .csv, can you give an example or hint how to match each image in the cxr dataset to the rows in the csv? sorry if it's a silly question

Einstellung changed discussion status to open Apr 20, 2023

Einstellung changed discussion status to closed Apr 29, 2023

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment