1. KRR Avatar KRR
  2. software
  3. IDP
  4. Issues

Generating functions <-> predicates

Issue #375 resolved
Broes De Cat created an issue

In e.g. dfasat, the generator when Trans is a partial function is much more complex than the generator for when Trans is a predicate. Trans is given in the structure. Can we improve this?

Comments (7)

  1. Bart Bogaerts

    Removed partial-function checks from GenerateBddAccToBounds

    For 3 reasons: * They were incorrect: only conditions about having an image were added, never about not having one (in the different cases true/false, other things should have happened) resulting in generating too much * They were not needed: out-of-bounds checks happen by the type checks above, other non-image checks happen by taking the inverse of the graphinter at appropriate places * They slowed down the grounding A LOT see issue #375. Fixes it at least partially, but I didn't compare with the predicate version yet.

    → <<cset 446365dbd247>>

  2. Bart Bogaerts

    Removed partial-function checks from GenerateBddAccToBounds

    For 3 reasons: * They were incorrect: only conditions about having an image were added, never about not having one (in the different cases true/false, other things should have happened) resulting in generating too much * They were not needed: out-of-bounds checks happen by the type checks above, other non-image checks happen by taking the inverse of the graphinter at appropriate places * They slowed down the grounding A LOT see issue #375. Fixes it at least partially, but I didn't compare with the predicate version yet.

    → <<cset f32cd1a9e798>>

  3. Bart Bogaerts

    Removed partial-function checks from GenerateBddAccToBounds

    For 3 reasons: * They were incorrect: only conditions about having an image were added, never about not having one (in the different cases true/false, other things should have happened) resulting in generating too much * They were not needed: out-of-bounds checks happen by the type checks above, other non-image checks happen by taking the inverse of the graphinter at appropriate places * They slowed down the grounding A LOT see issue #375. Fixes it at least partially, but I didn't compare with the predicate version yet.

    → <<cset eecc61e04e12>>

  8. Log in to comment
Assignee
Type
bug
Priority
major
Status
resolved
Component
–
Version
–
Votes
0
Watchers
2
Jira: the preferred issue tracker for Bitbucket. Join the team!