lieb_ando: use partially specified problem to get correct vexity

[ericphanson]

This is in fact the approach used in CVXQUAD (https://github.com/hfawzi/cvxquad/blob/master/lieb_ando.m), but we didn't have those yet at the time.

This way we don't have to create an atom! Pretty convenient.

Also: I noticed in many of the CVXQUAD-derived atoms, we expect some variables to be PSD, but don't check or enforce it. Could be something we might want to do.

closes #683

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.20%. Comparing base (2978c32) to head (9441aef).
Report is 1 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #684   +/-   ##
=======================================
  Coverage   98.19%   98.20%           
=======================================
  Files          89       89           
  Lines        5162     5168    +6     
=======================================
+ Hits         5069     5075    +6     
  Misses         93       93           

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

[ericphanson]

unexpected/unrelated docs failure:

ERROR: LoadError: LoadError: ArgumentError: Package TimeZones does not have RecipesBase in its dependencies:
- If you have TimeZones checked out for development and have
  added RecipesBase as a dependency but haven't updated your primary
  environment's manifest file, try `Pkg.resolve()`.
- Otherwise you may need to report an issue with TimeZones

[odow]

I just restarted it. The TimeZones release is only 2 hours old: https://github.com/JuliaTime/TimeZones.jl/releases/tag/v1.16.0, so I wonder if it's an issue with the cache

[odow]

Ha. I just arrived at JuliaTime/TimeZones.jl#459 (comment) to find you there too.

As a work-around, we can update the docs to build on Julia 1.10

[ericphanson]

yeah, that works. I put up a patch also: JuliaTime/TimeZones.jl#461