Repository navigation
Fix up comparisons with disperate variable sets #11
Description
When a theory doesn't mention all of the variables, then counting, comparing, etc all become incorrect. Ultimately, a fix will need to come in the form of an NNF potentially storing a superset of the variables currently mentioned in descendants via the walk functionality.
Activity
haz commentedon Jul 30, 2020
This assumption should go away once this issue is resolved.
blyxxyz commentedon Jul 30, 2020
I gave this some thought during initial development but didn't come up with anything that really felt right.
One fix could be to add an optional keyword argument
namesfor passing an explicit set of variables.Comparing using the current behavior of
.equivalent()is resistant to this problem, at least.haz commentedon Jul 30, 2020
Ya, I considered the
namesargument too. Problem there is that it proliferates -- many places might need it (or have sub-procedures that need it). Themake_smoothoperation is certainly one that would require it...