rewrite math in code to match paper

[chriscoey]

fixes partially #368, but the failure case tests are commented out rn because they do not pass with GLPK

brings the math in the code in line with the math in the MPC paper

[mlubin]

Agreed that this will make comparing the code and the paper much easier

[codecov[bot]]

Codecov Report

Merging #377 into master will increase coverage by 0.28%. The diff coverage is 96.11%.

@@            Coverage Diff             @@
##           master     #377      +/-   ##
==========================================
+ Coverage   91.74%   92.02%   +0.28%     
==========================================
  Files           4        4              
  Lines        1696     1744      +48     
==========================================
+ Hits         1556     1605      +49     
+ Misses        140      139       -1

Impacted Files	Coverage Δ
src/conic_algorithm.jl	92.58% <96.11%> (+0.38%)	:arrow_up:

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 64f274e...709b1b6. Read the comment docs.

[codecov[bot]]

Codecov Report

Merging #377 into master will decrease coverage by 0.48%. The diff coverage is 94.9%.

@@            Coverage Diff             @@
##           master     #377      +/-   ##
==========================================
- Coverage   91.74%   91.25%   -0.49%     
==========================================
  Files           4        4              
  Lines        1696     1739      +43     
==========================================
+ Hits         1556     1587      +31     
- Misses        140      152      +12

Impacted Files	Coverage Δ
src/solver.jl	80.39% <ø> (ø)	:arrow_up:
src/conic_algorithm.jl	91.49% <94.9%> (-0.71%)	:arrow_down:

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 64f274e...366b1ba. Read the comment docs.

[chriscoey]

the new exp cone s=0 failure case tests are a bit fragile on GLPK unfortunately. CPLEX is fine

i will comment them for now and leave the issue open. we may need to play with tolerances. there is also a potential issue with r = 0