Support JuMP.value for Logical Variables

[pulsipher]

Closes #94. It also alphabetizes the Project file to fix an Aqua test failure.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (3879754) 99.90% compared to head (97de4cd) 99.90%.

Additional details and impacted files

@@           Coverage Diff           @@
##           master      #96   +/-   ##
=======================================
  Coverage   99.90%   99.90%           
=======================================
  Files          10       10           
  Lines        1001     1003    +2     
=======================================
+ Hits         1000     1002    +2     
  Misses          1        1           

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

[hdavid16]

Thanks for putting this together. What are your thoughts on returning the corresponding Bool instead of 0 or 1? Should be a simple update: value(binary) == 1 (or use isone).

[pulsipher]

Thanks for putting this together. What are your thoughts on returning the corresponding Bool instead of 0 or 1? Should be a simple update: value(binary) == 1 (or use isone).

I thought about this, but the problem is MIP solvers return binary variable solutions as Float64 not 0 or 1. So, we would need to use some sort of tolerance to round to the appropriate integer solution and then convert to Bool. I would prefer to not hard-code some tolerance or add more complexity to value by adding a keyword that would only work on logical variables.

[hdavid16]

That is a good point. Since the value is binary, what about comparing it to 0.5 (value > 0.5)?

[pulsipher]

Ok, I have changed it to round based on 0.5.