DM-48045: Create template metrics and template QC process for Prompt Processing. #408

erinleighh · 2025-06-30T16:23:49Z

No description provided.

jrmullaney

Code looks good, but I've got a few thoughts on how we structure this. Mostly along the lines of "can we combine this with that"?

I'm willing to be persuaded either way, so don't feel I'm saying we have to go one way or the other. Let me know your thoughts on what I've said.

Oh, and you'll probably have seen that it's failing GitHub checks on black formatting and isort.

jrmullaney · 2025-07-01T09:32:12Z

pipelines/deepCoaddQualityCore.yaml

@@ -0,0 +1,22 @@
+description: |


Do we really want to add another two pipelines on top of coaddQualityCore.yaml?

Could we instead just add this to coaddQualityCore.yaml, using either deep or template as the default connections.coaddType, then override that in either drp_pipe or ap_pipe (depending on which we choose - deep or template).

I think I'd prefer for the default here to be deep, since that's what's generally found in coaddQualityCore.yaml. We'd just make sure that it wasn't run as part of drp_pipe (if we didn't want it to be).

I saw the chat on dm-science-pipelines about overriding coaddName, but I don't think that's referring to this - I think that's specifically referring to doing all the coaddQualityCore tasks with template as well as deep.

jrmullaney · 2025-07-01T09:33:11Z

pipelines/templateQualityCore.yaml

@@ -0,0 +1,22 @@
+description: |


See above comment.

jrmullaney · 2025-07-01T10:24:28Z

python/lsst/analysis/tools/actions/plot/coaddDepthPlot.py

+
+        plt.subplots_adjust(hspace=0, wspace=0)
+
+        patch_counter = 90  # The top left corner of a tract is patch 90


Isn't hard-coding this risky? Just thinking that other skymaps may not have the same number of patches per tract.

This class should be using the skymap tractInfo. See PerTractPropertyMapPlot for an example.

jrmullaney · 2025-07-01T10:40:05Z

python/lsst/analysis/tools/actions/vector/vectorActions.py

        yield self.key_is_target_star, Vector
+
+
+class UniqueAction(VectorAction):


Is this used? I can't find anywhere in the PR where this is used. Was it added, then not needed, or have I missed something?

It's not used and I ended up removing it from the PR. :)

jrmullaney · 2025-07-01T10:52:08Z

python/lsst/analysis/tools/actions/vector/selectors.py

+
+    def __call__(self, data: KeyedData, **kwargs) -> Vector:
+        patches: Optional[tuple[str, ...]]
+        match kwargs:


While I like what this is doing grammatically (patch vs. patches 😄 ), and I see that we do the same for band/bands, I wonder whether we should just keep it simple and have just patches=None as a keyword so that it's easier to see at a glance.

Dan also added a PatchSelector so I deferred to his and removed mine from the PR. :)

jrmullaney · 2025-07-01T10:57:32Z

python/lsst/analysis/tools/atools/coaddInputCount.py

+            self.produce.metric.units[f"{name}_stdev"] = ""
+
+
+class CoaddQualityPlot(AnalysisTool):


Why not combine the metric and the plot generation into a single atool? Or do you want to keep them separate so that plots are only generated when needed, but metrics are made for everything?

The plots can take a while to generate depending on how much data so I kept 'em separate. When I run the QC metrics afterburner pipeline, I skip the plot generation unless it's specifically asked for.

jrmullaney · 2025-07-01T11:14:18Z

python/lsst/analysis/tools/tasks/coaddDepthSummaryPlot.py

+
+__all__ = (
+    "CoaddDepthSummaryPlotConfig",
+    "CoaddDepthSummaryPlotTask",


Similarly, do we want to combine these two tasks into one? Or keep them separate so that we don't have to make plots if all we want are metrics?

taranu · 2025-07-07T16:26:03Z

python/lsst/analysis/tools/actions/vector/selectors.py

        return cast(Vector, mask)


+class PatchSelector(VectorAction):


FYI I added a PatchSelector recently, but it doesn't look like you actually need it.

taranu · 2025-07-07T16:27:47Z

python/lsst/analysis/tools/actions/plot/coaddDepthPlot.py

+        while patch_counter >= 0:
+            for n in range(10):  # column index
+                ax = plt.subplot(10, 10, m + 1)  # there are 10x10 patches per tract
+                patchSelector = PatchSelector(vectorKey='patch', patches=[patch_counter])


There's no point making a ConfigurableAction inside a call method where it can't actually be reconfigured in a pipeline. You already put patch and band in the schema so you may as well just do patch_mask = data["patch"] == patch_counter, and same for band_mask.

mrawls · 2025-07-19T05:32:32Z

Some unsolicited suggestions re. pipelines

I agree with James that having deepCoaddQualityCore.yaml and also template[Coadd?]QualityCore.yaml is kind of annoyingly overly specific when the only real connection the pipeline needs to function is... an n_image that has either the noun "deep" or "template" attached to it.
I don't think any of these pipelines belong in coaddQualityCore. Consider a totally different name to better disentangle it from "quality" and "core"... maybe coaddDepthMetrics.yaml? ... and then have it configurable - more or less as James says - to take "deep" or "template." The key distinction here is that NO object tables are involved whatsoever, only n_images. Therefore, I pretty strongly think the appropriate default noun should be "template." Because "deep" coadds have tables alongside them to metric-ify forever, and "template" coadds do not.
Adding it to both drp_pipe and ap_pipe for template_coadds is something that should happen, but on a separate ticket.

jrmullaney

Looks good.

I've left a couple of comments about naming outputs, and a request to add an example figure for your new plot action (I've only been made aware of these example figures since I last reviewed this ticket).

Approved.

jrmullaney · 2025-10-01T14:51:11Z

python/lsst/analysis/tools/actions/plot/coaddDepthPlot.py

+        -------
+        fig : `~matplotlib.figure.Figure`
+            The resulting figure.
+        """


Could you add an example figure here, please? Sophie Reed pointed this out to me when I recently made a new plot action. An example is here (but most other plot actions have them, too). It took me a while to figure out where I needed to put the plot - it goes in doc/_static/analysis_tools/ in this repo, then the doc generator finds it when it builds the docs for the website. Pretty nifty!

jrmullaney · 2025-10-01T14:57:14Z

python/lsst/analysis/tools/atools/coaddInputCount.py

 class CoaddInputCount(AnalysisTool):
-    """skyPlot and associated metrics indicating the number
-    of exposures that have gone into creating a coadd.
+    """Coadd-wide metrics pertaining to how many exposures have gone into


Could you change this to "Tract-wide" instead, to make it clear that it's on a per-tract basis, rather than a per-patch basis (at least, I think that's the case)?

I don't actually know much about this metric. The comment change was mostly to alert folks that this does not work for template_coadds. I'll put the comment changes into their own commit and then ask for clarity on Slack. Thanks for catching that!

jrmullaney · 2025-10-01T15:11:44Z

python/lsst/analysis/tools/tasks/coaddDepthSummaryAnalysis.py

+    )
+
+
+class CoaddDepthSummaryAnalysisTask(AnalysisPipelineTask):


Does this need to be an AnalysisPipelineTask? It seems it should be a just a regular PipelineTask. Does it use any of the AnalysisPipelineTask infrastructure?

jrmullaney · 2025-10-01T15:18:24Z

python/lsst/analysis/tools/tasks/coaddDepthSummaryAnalysis.py

+    dimensions=("tract", "skymap"),
+    defaultTemplates={
+        "coaddType": "",  # set as either deep or template in the pipeline
+        "outputName": "coadd_depth_table",


Do you want the outputName to include the type of coadd?

I would love to figure out how to do that! I couldn't manage any sort of "outputName": "{coaddType}_coadd_depth_table" that actually worked. If you know the trick, I'd love to update these. :)

For this specific task, you can just turn it into a regular PipelineTask, in which case you can just do:

statTable = cT.Output( doc="Table with resulting n_image based depth statistics.", name="{coaddType}_coadd_depth_table", storageClass="ArrowAstropy", dimensions=("tract", "skymap"), )

It doesn't need to be an AnalysisPipelineTask. You can then get rid of outputName in the defaultTemplates.

jrmullaney · 2025-10-01T15:27:55Z

python/lsst/analysis/tools/tasks/coaddDepthSummaryPlot.py

+    dimensions=("tract", "skymap"),
+    defaultTemplates={
+        "coaddType": "",
+        "outputName": "n_image",


Similarly, should this include the coadd type?

jrmullaney · 2025-10-01T15:28:14Z

python/lsst/analysis/tools/tasks/coaddDepthTableTractAnalysis.py

+class CoaddDepthTableTractAnalysisConnections(
+    AnalysisBaseConnections,
+    dimensions=("tract", "skymap"),
+    defaultTemplates={"inputName": "coadd_depth_table", "outputName": "coadd_depth_summary"},


Should this include the coadd type?

erinleighh requested a review from jrmullaney June 30, 2025 17:07

jrmullaney reviewed Jul 1, 2025

View reviewed changes

taranu reviewed Jul 7, 2025

View reviewed changes

erinleighh force-pushed the tickets/DM-48045 branch 5 times, most recently from 0748c64 to 1948c9c Compare July 19, 2025 00:41

erinleighh force-pushed the tickets/DM-48045 branch 2 times, most recently from 91bddf5 to e277a29 Compare August 27, 2025 17:20

erinleighh force-pushed the tickets/DM-48045 branch 6 times, most recently from 3669189 to 36baf53 Compare September 10, 2025 21:52

erinleighh force-pushed the tickets/DM-48045 branch 4 times, most recently from cbab6b9 to d935f0a Compare September 30, 2025 16:46

jrmullaney approved these changes Oct 1, 2025

View reviewed changes

erinleighh force-pushed the tickets/DM-48045 branch 2 times, most recently from f9afff1 to 48d2fb4 Compare October 8, 2025 19:51

erinleighh added 4 commits October 15, 2025 14:15

Add plot action for coadd depth.

4d1a6e7

Add tasks and pipeline for template metrics.

d5fce5c

Add configs to allow for template coverage maps.

2d968dd

Update docstring and comments on CoaddInputCount.

bfa3cb5

erinleighh force-pushed the tickets/DM-48045 branch from 48d2fb4 to bfa3cb5 Compare October 15, 2025 21:19

erinleighh merged commit 6e3f52f into main Oct 16, 2025
16 checks passed

erinleighh deleted the tickets/DM-48045 branch October 16, 2025 03:08


		plt.subplots_adjust(hspace=0, wspace=0)

		patch_counter = 90 # The top left corner of a tract is patch 90

		yield self.key_is_target_star, Vector


		class UniqueAction(VectorAction):

		self.produce.metric.units[f"{name}_stdev"] = ""


		class CoaddQualityPlot(AnalysisTool):

DM-48045: Create template metrics and template QC process for Prompt Processing. #408

DM-48045: Create template metrics and template QC process for Prompt Processing. #408

Uh oh!

Conversation

erinleighh commented Jun 30, 2025

Uh oh!

jrmullaney left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mrawls commented Jul 19, 2025

Uh oh!

jrmullaney left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

erinleighh Oct 2, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jrmullaney Oct 2, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

jrmullaney left a comment •

edited

Loading

erinleighh Oct 2, 2025 •

edited

Loading

jrmullaney Oct 2, 2025 •

edited

Loading