now keeping tile, petal, and loc #1060
Conversation
|
@julienguy , this should fix your complaint the other day about not having the tile information in the delta files. Can you review that this is indeed what you needed? |
| @@ -338,6 +338,12 @@ def format_data(self, | |||
| "z": row['Z'], | |||
| } | |||
| args["log_lambda"] = np.log10(spec['WAVELENGTH']) | |||
| if "TILEID" in row: | |||
| args["tile"] = row['TILE'] | |||
There was a problem hiding this comment.
Yes, I have fixed the bug (fixed it in a commit). Should we rename the tile member in DesiForest to tileid as well?
There was a problem hiding this comment.
maybe that would be good for consistency... Esp given that e.g. eBOSS did have a PLATE instead of TILEID anyway, so it's a DESI specific thing.
| if "PETAL_LOC" in row: | ||
| args["petal"] = row['PETAL_LOC'] | ||
| if "NIGHT" in row: | ||
| args["night"] = row['NIGHT'] | ||
|
|
There was a problem hiding this comment.
Should we add e.g. FIBER or EXPID? Are these fields all read earlier from HDU2 (EXP_FIBERMAP) of the coadd-files, or just when reading spectra-files?
There was a problem hiding this comment.
Just checked and here we'd just copy over info from the catalogue, is all this info in the actual quasar cats? or would we need to assemble the data from the coadd-files?
There was a problem hiding this comment.
possibly, the FIBER would be nice to have. I'm not sure about the EXPID. @julienguy , what do you think?
There was a problem hiding this comment.
I just ran a test with this branch and this is just not working. Nothing is saved for NIGHT PETAL TILEID in the METADATA of the delta files.
Example run, ini file:
/global/cfs/cdirs/desicollab/users/jguy/lya/instrumental-systematics/deltas/delta-civ-tile-2-sub-1/picca_delta.ini
And delta files in :
/global/cfs/cdirs/desicollab/users/jguy/lya/instrumental-systematics/deltas/delta-civ-tile-2-sub-1/Delta
|
In case we still want to add this feature:
|
|
@iprafols or @calumgordon if I remember from the latest Lya soft telecon, you wanted to rework this PR, any news ? |
|
I guess given this PR is based on an out-of-date master with merge conflicts, it might be easier to start from scratch. |
This should propagate the read values of tile, petal and night values to the delta files, if they are available in the catalogue. They are stored as a string. Lines of sight with multiple observations contain all the values separated by "-".