DM-40186: Update dynamic detection task defaults and logging #349
Conversation
| @@ -123,6 +123,12 @@ def calculateThreshold(self, exposure, seed, sigma=None): | |||
| sigma : `float`, optional | |||
| Gaussian sigma of smoothing kernel; if not provided, | |||
| will be deduced from the exposure's PSF. | |||
| minFractionSourcesFactor : `float` | |||
| Change the fraction of required sky sources from that set in | |||
| ``self.config.minFractionSources`` by this factor. | |||
There was a problem hiding this comment.
Might want to document here that this is intended for use in the background-tweak pass (even though it's just one method for both passes), since using this instead of just modifying minFractionSources in the first threshold-setting pass would just be confusing (and I was at first quite confused to find the option here, given the method's name).
| f"{nGoodPix} ({100*nGoodPix/nPix:.1f}%) in this patch, so there should be " | ||
| "sufficient area to locate suitable sky sources. Something is amiss. " | ||
| "(Note that the fraction of \"good\" pixels that have the DETECTED or " | ||
| f"DETECTED_NEGATIVE mask bit set is {100*nDetectedPix/nGoodPix:.1f}%).") |
There was a problem hiding this comment.
Propose rewording this:
However, {nGoodPix}/{nPix} pixels are not marked NO_DATA or BAD, so there should be sufficient area to locate suitable sky sources. Note that {nDetectedPix} were marked as DETECTED or DETECTED_NEGATIVE.
It was the "However there are Number (%s)" beginning in particular that didn't read well to me, but I also have a slight preference for putting the original numbers into the logs directly, even if that makes the reader compute some percentages themself.
laurenam
force-pushed
the
tickets/DM-40186
branch
from
a6582d4
to
5e12d00
Compare
As the number of input visits increases, the number of pixels with the INTRP mask bit set increases, which eventually leads to a very large fraction pixels with this mask bit set. Thus, if INTRP is included in the list of masks to avoid when laying down sky objects, very deep coadds will end up with little-to-no available locations for sky objects, leading to a failure in the dynamic detection. Thus, the safest thing to do is leave it out of the masks to avoid list.
In the dynamic detection task, compute and log the number of "good" pixels in the exposure, where "good" is defined simply as not having either the NO_DATA or BAD mask bit set. If this fraction is greater than 20% of the total number of pixels, yet we still fail to locate enough spots to lay down sky objects, raise an RuntimeError exception. The message includes the fraction of the "good" pixels marked either DETECTED or DETECTED_NEGATIVE as this is a likely reason for this situation.
laurenam
force-pushed
the
tickets/DM-40186
branch
from
5e12d00
to
48b8a4b
Compare
No description provided.