# Retrieving the raw WCS from metadata is buggy

XMLWordPrintable

## Details

• Type: Story
• Status: To Do
• Resolution: Unresolved
• Fix Version/s: None
• Component/s: None
• Labels:
None
• Team:
Data Release Production

## Description

In DM-19731, Jim Bosch noted "some interesting and unexpected things" when looking into initial HSC WCSs. It turns out that there is something going awry when fetching the raw WCS from metadata and then applying the distortion correction to it, as the following plots demonstrate:

Using

 rawWcs = butler.get("raw_wcs", dataId) 

they look like (this is the funky behavior noted in DM-19731):

However, using

 raw = butler.get("raw", dataId) rawWcs = raw.getWcs() 

they look as expected:

Jim Bosch suspects this might be related to the stripping ordering problem he ran into here: https://github.com/lsst/daf_butler/commit/ac5032c6361cfb3a6f8d1dd78cf2b52fb6baa3e3

This needs resolving and, for now, the raw WCS obtained from metadata should not be trusted.

## Activity

Hide
John Parejko added a comment -

I'm not sure that this actually needs to be fixed, if we go with the approach described in DM-19951, since it won't use the metadata at all. In that case, it does mean we need to fix HSC camera geometry, but that might be as simple as fixing the units.

You do bring up a point I hadn't considered, though, which is that I don't really understand how the raw_wcs butler datatype is constructed. The function I'm writing for DM-20154 is what should be called to do that, but I don't know how to make that happen.

Show
John Parejko added a comment - I'm not sure that this actually needs to be fixed, if we go with the approach described in DM-19951 , since it won't use the metadata at all. In that case, it does mean we need to fix HSC camera geometry, but that might be as simple as fixing the units. You do bring up a point I hadn't considered, though, which is that I don't really understand how the raw_wcs butler datatype is constructed. The function I'm writing for DM-20154 is what should be called to do that, but I don't know how to make that happen.
Hide
John Parejko added a comment -

Given RFC-616 and the implementations in DM-20154 and DM-20742, I think we can close this as Invalid, since it is about to become irrelevant.

Show
John Parejko added a comment - Given RFC-616 and the implementations in DM-20154 and DM-20742 , I think we can close this as Invalid, since it is about to become irrelevant.

## People

• Assignee:
Unassigned
Reporter:
Lauren MacArthur
Watchers:
Chris Morrison, Jim Bosch, John Parejko, John Swinbank, Lauren MacArthur