bart-workshop icon indicating copy to clipboard operation
bart-workshop copied to clipboard

data missing for 4th demo "wave-cs"

Open welcheb opened this issue 9 years ago • 8 comments

Thank you for the helpful demos of BART! I successfully ran the first 3 demos on Mac OS X, however, the data seems to be missing for the 4th demo "wave-cs".

welcheb avatar Jan 23 '16 04:01 welcheb

Thanks for the feedback. Github has a data limit size of 100 MB, so I could not upload the dataset here (~300MB). I'm still trying to find a way to upload the data so that people can get them easily.

For now, here's a google drive link containing the wave-cs data: https://drive.google.com/folderview?id=0BxOlBDnZVRv6WWVadGtxd2RwYlk&usp=sharing

frankong avatar Jan 23 '16 07:01 frankong

Frank,

How about MRIDATA.org http://mridata.org/ ?

On Jan 22, 2016, at 11:21 PM, Frank Ong [email protected] wrote:

Thanks for the feedback. Github has a data limit size of 100 MB, so I could not upload the dataset here (~300MB). I'm still trying to find a way to upload the data so that people can get them easily.

For now, here's a google drive link containing the wave-cs data: https://drive.google.com/folderview?id=0BxOlBDnZVRv6WWVadGtxd2RwYlk&usp=sharing https://drive.google.com/folderview?id=0BxOlBDnZVRv6WWVadGtxd2RwYlk&usp=sharing — Reply to this email directly or view it on GitHub https://github.com/mikgroup/bart-workshop/issues/1#issuecomment-174156628.

mlustig avatar Jan 23 '16 07:01 mlustig

That’s a great idea! I will check it out!

-Frank

On Jan 22, 2016, at 11:22 PM, Miki Lustig [email protected] wrote:

Frank,

How about MRIDATA.org http://mridata.org/ ?

On Jan 22, 2016, at 11:21 PM, Frank Ong [email protected] wrote:

Thanks for the feedback. Github has a data limit size of 100 MB, so I could not upload the dataset here (~300MB). I'm still trying to find a way to upload the data so that people can get them easily.

For now, here's a google drive link containing the wave-cs data: https://drive.google.com/folderview?id=0BxOlBDnZVRv6WWVadGtxd2RwYlk&usp=sharing https://drive.google.com/folderview?id=0BxOlBDnZVRv6WWVadGtxd2RwYlk&usp=sharing — Reply to this email directly or view it on GitHub https://github.com/mikgroup/bart-workshop/issues/1#issuecomment-174156628.

— Reply to this email directly or view it on GitHub https://github.com/mikgroup/bart-workshop/issues/1#issuecomment-174156684.

frankong avatar Jan 23 '16 07:01 frankong

Thank you! I downloaded the files from drive.google.com and the 4th demo ran successfully.

As far as posting data too large for github.com, we have taken two approaches:

  1. Post using services such as http://www.copy.com and include a download list and download script with the github repository. If we have to switch cloud services in the future, we can update the github repo to point to the updated download links. See for example the usage steps in https://github.com/welcheb/MRM_Sengupta_Moving_Table_Golden_Angle_MRI/blob/master/SETUP.md
  2. Post using http://www.zenodo.org, which is backed by the EU and created for CERN. I think zenodo will be around for a long time. Zenodo also creates a citable DOI for each entry. There is supposedly no size limit per post other than a maximum of 2GB per file. The downside is that download is very slow. We have started to redundantly post at zenodo. For example, the data for the golden angle moving table paper is available here: http://dx.doi.org/10.5281/zenodo.16747 Since BART already has zenodo entries, this is probably not news to your team!

The datasets at http://www.mridata.org look like a great resource for the MRI reconstruction community! Is there a plan for a way for outside users to post data there?

best regards, Brian

welcheb avatar Jan 23 '16 16:01 welcheb

I recently became aware of a very easy way to share large binaries (data, installers, etc.) using Github itself. For any release, an unlimited number of files up to 2GB each can be attached to the release. The download links are direct and be used with scripts, wget, curl, etc. to automate download. Github does not limit the total size of files nor the download bandwidth! See https://help.github.com/articles/distributing-large-binaries for more details.

You can also use the Github API to track how many times an "asset" has been downloaded. For example, the ISMRMRD team posted a data file with release v1.2.3-data (https://github.com/ismrmrd/ismrmrd/releases/tag/v1.2.3-data) that has been downloaded 30 times (https://api.github.com/repos/ismrmrd/ismrmrd/releases/assets/784648). You can get the asset list for a given release with an api URL like this https://api.github.com/repos/ismrmrd/ismrmrd/releases/tags/v1.2.3-data

welcheb avatar Mar 21 '16 16:03 welcheb

The database in the google drive link shared for the demo "wave-cs" seems to be missing . Could you please share it. Thanks.

kevinYitshak avatar Oct 17 '19 08:10 kevinYitshak

The data should be up now

frankong avatar Oct 17 '19 15:10 frankong

Thank you!

kevinYitshak avatar Oct 17 '19 17:10 kevinYitshak