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
[WIP] EDR cleanup #167
[WIP] EDR cleanup #167
Conversation
I added a unit test that I know will fail (at least for now), because these columns do not have types or descriptions:
None of these columns appear in any data model file, so it's not obvious where these come from. |
@aureliocarnero do you know where the mystery columns listed above come from? Were they possibly in an early version of the LSS catalog? |
Porting a message I just sent Ben on Slack: It's likely these columns may have been inherited from the If so:
|
Thank you @geordie666.
|
Note: I'm using the definition |
Based on a comment from Ben on Slack, I think the resolution is straightforward, here.
I don't think these columns will be included in the EDR. So, there's no need to include them in the data model. As I noted above, I suspect these quantities may have been blindly inherited from the |
Thank you @geordie666! |
Dear @weaverba137 et al. I think you found the origin to these columns. They are not used in LSS. Cheers |
Update: as of now there are no longer any "default" descriptions as described in #163. Instead, every remaining instance now has
to find missing information. At this point there is nothing special about coordinates or pm files (#120) in I believe I have also satisfied all the suggestions in #140, except where TODOs with paper links need to be filled in. That just leaves verification and remaining TODOs to work on. |
Mapping data model files to files that are actually in EDR:
|
None of the intermediate fiberassign files matches what is on disk for EDR. The directory path is significantly different. |
In the (non-intermediate) fiberassign files, the uncompressed fiberassign-EXPID.fits files have a lot of differences from the compressed fiberassign-EXPID.fits.gz files. But the data model for the compressed files is good. |
sorry, just checking: what do you mean by "a lot of differences"?
this folder is a bit "special", in the sense that we discovered that the
|
@araichoor, please be careful here. The uncompressed non-intermediate files have many differences from the compressed files. Those are the differences I am talking about. That said,
That is an issue with the directory path not the files. The intermediate files are described in this directory: https://github.com/desihub/desidatamodel/tree/main/doc/DESI_ROOT/survey/fiberassign/SURVEY/TILEXX. However, the pattern in EDR is:
And |
@araichoor, PS I am using this space to take notes on the many data model tasks left to perform before EDR, and the notes are very raw. They may even be edited after being written. I would not spend a lot of time reading this thread until the PR is actually ready for review. |
Every
|
…into edr-cleanup
* origin/edr-cleanup: resolve TODO items in DESI_SPECTRO_REDUX/SPECPROD
…into edr-cleanup
Merging this after discussion with @sbailey. Remaining issues can be settled with more targeted PRs. |
This PR: