# flag non-finite measurement uncertainties

XMLWordPrintable

## Details

• Type: Epic
• Status: To Do
• Resolution: Unresolved
• Fix Version/s: None
• Component/s: None
• Labels:
• Epic Name:
finite uncertainties

## Description

All of our measurement algorithms that produce _flag fields need to confirm that both non-finite values, and non-finite uncertainties are flagged with that _flag field. This may already be true for some measurements, but it is known not to be true for others (e.g. sdss_centroid).

We need to document this meaning of the _flag fields and document that this is a requirement of all new measurement algorithms.

We should include in data validation a check on finiteness of measurements and uncertainties of unflagged sources as a way to confirm that our flags are working correctly.

This is the implementation ticket for RFC-428.

## Activity

Hide
Gregory Dubois-Felsmann added a comment -

Is this still work that remains to do? Jim Bosch? I came to this through DM-6655, which I'm taking another look at.

Show
Gregory Dubois-Felsmann added a comment - Is this still work that remains to do? Jim Bosch ? I came to this through DM-6655 , which I'm taking another look at.
Hide
Jim Bosch added a comment -

We've been fixing violations of this as they come up (e.g. DM-22958), but I do not think we have any comprehensive test in CI.

Show
Jim Bosch added a comment - We've been fixing violations of this as they come up (e.g. DM-22958 ), but I do not think we have any comprehensive test in CI.

## People

• Assignee:
Unassigned
Reporter:
John Parejko
Watchers:
Gregory Dubois-Felsmann, Jim Bosch, John Parejko, John Swinbank