Disk margin calculations (python-control#1146)

* Initial version of disk margin calculation and example/test script

* Comment updates: update margins.py header, clarify import exception handler comment, fix typo in skew description of disk_margins docstring

* More work in progress on disk margin calculation, adding new prototype function to plot allowable gain/phase variations.

* Add disk_margin_plot to subroutine list in comment header in margins.py

* Follow-on to ba15789, add disk_margin_plot to list of functions within the margins module

* More work in progress on disk_margin_plot.  Corrected a typo/bug in the
calculation of 'f', the bounding complex curve.  Seems to look correct for balanced
(skew = 0) case, still verifying the skewed equivalent.

* Further progress/debugging on disk margin calculation + plot utility

* Clean up docstring/code for disk_margin_plot

* Clean up docstring/code for disk_margin_plot

* Remove debugging statements, update comments, add unit tests.

* Minor change to fix logic to find minimum across DGM, DPM numpy vectors

* Rename disk margin example, since unit tests are now written in control/tests/margin_test.py

* Remove unneeded dependencies from margins.py, used for debugging

* Minor updates to docstrings

* Undo d92fb20

* Minor tweaks to plots in example script for readability

* Fix typo in disk_margin_plot.

* Fix mag2db import hack/workaround and trim down disk_margin docstring.

* Add input handling to disk_margin, clean up column width/comments

* Move disk_margin_plot out of the library into the example script

* Recommended changes from the linter

* Follow-on to 5f34a7b

* Add disk_margins to function list

* Whittle down the docstring from disk_margins

* Put more comments in the disk margin example, add example to documentation

* Fixing docstrings

* Corrected expected values for 'no-slycot' condition in newly-added unit tests

* Attempt python-control#2 at 397efab, based on linter recommendation

* Address @murrayrm review comments.

* Update formatting per PEP8/@murrayrm review comments.  Add additional reference on disk/ellipse-based margin calculations.

* Follow-on to e8897f6: remove now-unnecessary import of importlib

* Update formatting per @murrayrm review comments

* Remove temporarily-added string from docstring

* Minor tweak to docstring to fit the word 'function' back into the description of skew = 0.0

recalculate loci for sisotool and rlocus on axis scaling