The code to do what is written in the description of the ticket—including modifications by Robert---can be found in: 

/home/plazas/jhome/WORK/DM-17490/FindDefects_may7.ipynb

It runs with w_2019_18 setup.  I attach a copy of the code as of today, for convenience. 

FindDefects_may7.py

(Changing my previous comment to say):

Given that this is not the final outcome of DM-17490, I think it is ok for this not to exist in a git repository in the short term.

Since paths are mutable, it's not safe to just refer to it by a filesystem location. However, attaching it to the ticket (as you have done) is fine for now.

Please make sure a current version of the code is attached when you are ready to send this ticket for review.

Robert and I have written code that does what is written in the description of this ticket, i.e., identifies n-sigma outliers from master darks and master flats. This code has been used with AuxTel, TS8, and comCam data. The code is in the following file: FindDefects.py

Documentation can be found in the doctrings inside the code, but the the function that does the main calculations is "find_hot_and_cold_pixels".  The code creates a directory ("output_directory") where ASCII files with the defects (as boxes) are printed. In addition, histograms and other diagnose plots are written to PDF files in that same directory.