Work on 16333 to develop gambit
Closed this issue · 10 comments
This will need to grab code from old-16333
and also include gambit as an optional package. This needs @theref and I to talk about it though.
Have got the parser documented and tested for LCP.
Have added a bunch of tests that now fail. Let me know if/when you're working on this. If I don't hear from you I will probably keep going with it.
No harm in adding more and/or checking if they're (as well as the docs around them) correct.
I've used # optional - gambit, lrs
at some point to say a test that should only run if both optional packages are installed. I'm not sure if that syntax is correct...
A part from potentially removing or at least refactoring the conversion of gambit output to integer where possible (not at all sure why this is in there, see my comment here: a16dc89 - perhaps work on features that we haven't discussed on branches so that we can talk about them...) this is more or less done. Two things are needed:
- Allowing init to take a gambit game
- Thorough re-read of everything
c402408 now has gambit fully integrated. It needs refactoring though as I'm importing gambit everywhere (and in ways that will crash if gambit is not installed).
Once that is fixed there are two remaining things before pushing to trac:
- Discuss parser (see comment here a16dc89)
- Thorough check and documentation (including moving part about running gambit in ipython etc...)
Also, the _gambit_game
method will need a couple of tests.
Have a branch before I start removing the gambit formatting: https://github.com/theref/sage-game-theory/tree/branch_formating_of_integer_output_of_gambit
This has removed the integer formatting: fa5f6e7
I think gambit is now being called in the best way possible (based on: https://groups.google.com/forum/#!topic/sage-devel/E4Qe15oRTbc). Here's the commit fa5f6e7. I expect that it might change after reviews...
Just a thorough read through left now :)
Just about to push this to trac!
Ok I've pushed a branch that corresponds to 16333 but is all rebased and stuff... https://github.com/theref/sage-game-theory/tree/rebased_version
I'm guessing that once reviews come in I'll work on that one... No idea what is going on.