Conversation
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## master #405 +/- ##
=======================================
Coverage 99.77% 99.77%
=======================================
Files 38 38
Lines 6586 6586
=======================================
Hits 6571 6571
Misses 15 15 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
|
Please get the docs to pass, so I can review |
* Change subheadings in variable.md + adjust InfiniteInterpolations func signature * Fix example output for InfiniteMathOptAI
…niteOpt.jl into resolveTutorial
pulsipher
left a comment
There was a problem hiding this comment.
See comments below. We need to be more careful with the math notation and I think we need to clearly define and discuss what the re-solve interface is at the beginning before diving into the example. Subsections for all the different definition steps is probably also unnecessary and distracts from the key focuses. The underlying use of transformation backends should also be made more clear since this is a key piece of information to contextualize what is happening.
pulsipher
left a comment
There was a problem hiding this comment.
Almost there, my comments are mostly minor now. The major item left is improving the diagram.
|
Also, is there a way to avoid using DifferentialEquations in this tutorial? Including it, increases the documentation build time to 15 minutes (up from 7 minutes). Since we only do one re-solve, perhaps we can just provide manual measurements and say one could use DifferentialEquations, but we don't run it in the documentation to minimize precompilation time. |
pulsipher
left a comment
There was a problem hiding this comment.
This is looking great now, just one final tweak and it'll be there.
pulsipher
left a comment
There was a problem hiding this comment.
Ok, another minor thing would be that we should not display the solver output, because this may change over time with different versions which will be annoying to maintain. Let's set it to silent and instead make generic queries. You can query the number of iterations via barrier_iterations(model)
This PR aims to add a comprehensive tutorial on how to setup efficient resolves on CPU & GPU.