Re-uploading Archived Datasets

Shared with the CNT

This procedure runs on infrastructure the lab and the CNT operate together. The CNT manual keeps its own copy; changes to the shared steps are agreed with the CNT.

What this page tells you

To fix an already-archived dataset, ask the data research coordinator to unarchive it to /project/eeg_process/azure_archive/recently_unarchived/HUPXXX, recreate the HUPXXX folder in cnt-fs with channel mapping, natusdir, and data collection from ieeg_metadata, cp the recordings back, then rerun the Processing for ieeg.org SOP from the mef conversion.

If there is an error in an uploaded dataset that has already been archived from cnt-fs, follow the steps below.

1. Ask the data research coordinator to put in a ticket to request that it be unarchived
2. He will move it to cnt1 in project/eeg_process/azure_archive/HUPXXX
3. Create a HUPXXX folder in cnt-fs in eeg_raw/ieeg_raw
   1. move the channel mapping, natusdir, and data collection files from ieeg_metadata to this HUPXXX folder
4. In Mobaxterm in the VDI, enter the unarchived HUPXXX folder in cnt1: cd /project/eeg_process/azure_archive/recently_unarchived/HUPXXX
5. Now you are going to copy the unarchived data from cnt1 into cnt-fs : cp -r Intracranial_EEG /mnt/cnt-fs/eeg_raw/ieeg_raw/HUPXXX
6. This will take some time to finish
7. Remember you need to following files to reupload:
   1. All the eeg recordings
   2. natusdir
   3. channel mapping
   4. data collection
8. Once you have everything unarchived, you can proceed with the Processing for ieeg.org (Processing for ieeg.org (natus2mef → validate → upload)) SOP, starting from the beginning with converting the recordings to mefs