DM-36208: Add a README to analysis_tools

[leeskelvin]

No description provided.

[parejkoj]

A couple of suggestions. Thanks for expanding this.

[parejkoj]

Suggested change

	The Analysis Tools package is designed to assist in the creation of quality assurance (QA) plots and metrics from the outputs of a data reduction pipeline.
	The Analysis Tools package is designed to assist in the creation of quality assurance (QA) plots and metrics from the outputs of a PipelineTask.

[leeskelvin]

Use of the phrase "data reduction pipeline" here was specific, as data derived from more than just a single pipeline task may be used to generate plots and metrics. I can't think of a neater way to summarize that than "data reduction pipeline".

[parejkoj]

"data reduction pipeline" is either too specific (not just DRP!) or not specific enough (it has to be one of our LSST Pipelines). Maybe "from the outputs of an LSST Pipeline" and link the latter two to https://pipelines.lsst.io/modules/lsst.pipe.base/creating-a-pipeline.html? I don't know if we have a better summary page for what we mean when we say "Pipeline", but it's a more specific thing than what non-LSST people mean.

[leeskelvin]

DRP != Data Reduction Pipeline.
DRP = Data Release Production.

Does that change your thinking? I think the phrase "data reduction pipeline" is sufficiently instantly recognisable as to what we mean here - namely, a pipeline that has reduced data.

Happy to add a URL link.

[parejkoj]

The link helps. We are specifically talking about our capital-P Pipelines, and that makes it clear.

[parejkoj]

"In this sense"? Maybe something like this:

Suggested change

	In this sense, a series of consistent, repeatable and high-quality plots and metrics can be generated for any given dataset.
	Analysis Tools allows users to generate consistent, repeatable, and high-quality plots and metrics on the output of a PipelineTask, no matter which dataset it is run on.

[leeskelvin]

Barring the pipeline task issue above, this is 👍

[parejkoj]

I don't think the "therefore" is necessary.

Suggested change

	These analysis tools therefore allow users to generate a series of consistent, repeatable and high-quality plots and metrics, regardless of which dataset they are run on.
	Analysis tools allows users to generate a series of consistent, repeatable and high-quality plots and metrics, regardless of which dataset they are run on.

[leeskelvin]

It helps if you read the fully rendered paragraph, I find: https://github.com/lsst/analysis_tools/tree/tickets/DM-36208.

If sentence 1 in the paragraph begins "The Analysis Tools package..." and then sentence 3 begins "Analysis tools...", it's too much repetition of the phrase in quick succession. This is why I initially preferred "In that sense...", as it's clear were still discussing in the context of sentence 1.

[leeskelvin]

Use of "therefore" is an attempt to break up that repetition in an attempt to satisfy your comment.

[leeskelvin]

PS - as Analysis Tools isn't yet linked in the static docs, I've added an additional sentence linking to the current daily build. I think this will be useful in the coming weeks, as in-kind collaborators begin adding code and wanting to reference what we have right now.

[parejkoj]

Good point on linking to the daily docs build: I think that's probably a good idea generally (at least until we have an official data release), since the docs are evolving pretty fast right now.