Add filtering for center packet loss

[PyxieLouStar]

Provide the option for flagging shapes only if the computed significance value is negative.

Description

Certain effects, such as correlator packet loss, show up as a negative in the SSINS. This change allows users to set the sig_thresh value for any shape as a negative number in order to have the shape only be flagged when the significance is negative. The behavior for positive sig_thresh is unchanged: the absolute value of the significance will be used when checking for the shape.

Checklist:

Main Code Body Checklist:

• Add or update docstring related to code change
• Add a unit test that verifies the efficacy of the code
• Write a tutorial if the feature would be useful to others to use
• Update CHANGELOG.md

Scripts Checklist:

• Add useful help strings for any arguments that are parsed
• Manually check that the script runs and gives proper results
• Update CHANGELOG.md

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (f87e430) 100.00% compared to head (5d2a49c) 100.00%.

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff            @@
##            master      #123   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            8         8           
  Lines          908       918   +10     
=========================================
+ Hits           908       918   +10     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[PyxieLouStar]

Negative sig_thresh for the narrow shape will not be allowed in this pr

[mwilensky768]

Hi @PyxieLouStar I looked at the changes and they look very straight-forward. The unit tests make sense.

I may revisit this code later since these if statements can technically be taken out of the loop (the answer is the same in every iteration of the loop so you can just precompute which branch to go down for each shape). I imagine this is a teeny tiny effect that will save us microseconds per obs so I'm just going to approve/merge.