Add documentation for outlier_detection

[stscijgbot-jp]

Issue JP-3342 was created on JIRA by Susan Kassin:

The ReadTheDocs page for contains minimal information on how outlier_rejection is performed.

It is difficult to follow the step arguments page for [outlier detection|#step-arguments-for-ifu-data]] without a description of what outlier_detection does.

[stscijgbot-jp]

Comment by David Law on JIRA:

I think this is already taken care of in the latest update: https://jwst-pipeline.readthedocs.io/en/latest/jwst/outlier_detection/outlier_detection_ifu.html

[stscijgbot-jp]

Comment by Susan Kassin on JIRA:

Thank you David Law!  The page in the latest update for "Outlier Detection for Slit-like Spectroscopic Data" needs a nice description like the one you linked to for the IFU.  The description on this page is more of a laundry list of what the algorithm does, and therefore is hard to follow.  It would be easier to follow if it were preceded by a summary of the algorithm.  https://jwst-pipeline.readthedocs.io/en/latest/jwst/outlier_detection/outlier_detection_spec.html

[stscijgbot-jp]

Comment by Jane Morrison on JIRA:

A nice picture of slit to stack of slits on output frame to  median image back to blotted frame I think would help explain what is going on. 

[stscijgbot-jp]

Comment by Howard Bushouse on JIRA:

Yes, I agree that the overall collection of pages within the entire step docs could be better organized. Part of the problem stems from the fact that there used to be just one fundamental, low-level algorithm used to actually detect outliers across stacks of multiple inputs, but each different mode (imaging, slit-like spectroscopy, IFU spectroscopy, TSO imaging and spectroscopy) had their own high-level wrappers around that base algorithm in order to manipulate the input data in different ways before then using the basic low-level algorithm to actually do the detection. That basic algorithm, which stems from the original application to simple imaging, is described here: https://jwst-pipeline.readthedocs.io/en/latest/jwst/outlier_detection/outlier_detection.html#outlier-detection-imaging . Each mode then has its own doc page describing how it's handled a little differently. And now that the IFU algorithm has been completely revamped and no longer uses the base algorithm, it has its own unique description. So yeah, it can be difficult to sort out. I think a complete reorganization of the pages within the step docs is in order. Right now it's laid out a little too much like a set of elaborate doc strings for each code module, rather than a real description for users.

[stscijgbot-jp]

Comment by Howard Bushouse on JIRA:

Similar issues with the docs are noted in JP-1647.

[stscijgbot-jp]

Comment by Rosa Diaz on JIRA:

I do agree that the outlier detection step documentation in read-the-docs is formatted completely different to the rest of the steps. The order of the information is different and it is difficult to follow the logic. Also, several list are not bulleted, which also make it difficult to follow too.

Because of this, I was not sure where the changes in ticket JP-3207 were in this documentation.

[stscijgbot-jp]

Comment by David Law on JIRA:

Adding a priority associating this with a general refactor of the outlier detection step.

[stscijgbot-jp]

Comment by Ned Molter on JIRA:

Fixed by #8880 

[stscijgbot-jp]

Comment by David Law on JIRA:

Reorganization looks great; closing.