Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • Apr 01 14:04
    quicquid commented #740
  • Mar 30 18:32
    Jenkins nightly build failure
  • Mar 30 18:04
    Jenkins develop failure
  • Mar 30 17:56
    jvierling commented #740
  • Mar 30 17:53

    jvierling on develop

    update iProver output parser disable superposition and non e… (compare)

  • Mar 30 13:35

    jvierling on prooftransform

    (compare)

  • Mar 30 13:35

    jvierling on structural_induction

    (compare)

  • Mar 30 13:34

    jvierling on tip-smt-parser

    (compare)

  • Mar 30 13:34

    jvierling on release-2.15

    (compare)

  • Mar 30 13:33

    jvierling on dls-algorithm-for-formula-equations

    (compare)

  • Mar 30 09:14
    Travis gapt/gapt (v2.15) passed (4067)
  • Mar 30 08:57

    jvierling on develop

    update release notes. Release 2.15. Bump version to 2.16-SNAPSHOT (compare)

  • Mar 30 08:44

    jvierling on v2.15

    (compare)

  • Mar 30 08:26

    jvierling on release-2.15

    Release 2.15. (compare)

  • Mar 30 08:22
    Travis gapt/gapt@fc1dcd3 (release-2.15) passed (4064)
  • Mar 30 07:50

    jvierling on release-2.15

    update release notes. (compare)

  • Mar 23 11:10

    shetzl on develop

    remove unnecessary includes (compare)

  • Mar 09 14:03

    jvierling on develop

    implement simplified DLS algori… refactor solveFormulaEquation return substitution from solveF… and 73 more (compare)

  • Mar 09 14:03
    jvierling closed #741
  • Mar 09 13:53
    Travis gapt/gapt (dls-algorithm-for-formula-equations) fixed (4060)
Andreas H. From
@andreasfrom
@jannikvierling nice, thank you. Most of them are malformed datatypes declaration exceptions and some are malformed expression. I'm running on the problems from the testing/TIP/tip-benchmarks-9c0088b73457ffbdb79159d487c9db2deb650381.tar.gz file in the repository.
Stefan Hetzl
@shetzl
@andreasfrom : Thanks for running the tests!
It is surprising to me that spind is significantly worse that analytic_sequential and analytic_indepedent.
It should be possible to be as good as / better than these without having to invest too much effort.
@jannikvierling Do you have some documentation about the analytic induction axioms that you could send Andreas?
@andreasfrom what does return status "failed" mean? I assume that "ok" means a proof has been found.
Jannik Vierling
@jvierling
@andreasfrom I will send you some notes I took about the independent and sequential induction axioms.
Andreas H. From
@andreasfrom
@jannikvierling Failed means saturated without finding a proof. So I guess it maybe doesn't always make sense to call it failed actually.
Jannik Vierling
@jvierling
@andreasfrom Have you tried to import the problems with 'TipSmtImporter.fixupAndLoad'. This uses the external tip-tools in order to do some preprocessing (monomorphisation, replacing int by nat, etc) on the problems.
Andreas H. From
@andreasfrom
@jannikvierling Yes, sorry I should have said I did try that, but it didn't seem to make much of a difference, giving me errors directly from the tip tool instead. Here is an example where fixupAndLoad just gives a different error (because of this, I think):
scala> val problem = TipSmtImporter.fixupAndLoad("benchmarks/isaplanner/prop_02.smt2")
java.util.NoSuchElementException: None.get
Jannik Vierling
@jvierling
Our Tip importer not very sophisticated so for the moment we will have to deal with certain cases where the error messages are not very comprehensible. However when using 'fixupAndLoad' many more problems can be parsed.
Andreas H. From
@andreasfrom
Okay, I will use that one instead. Thanks!
Jannik Vierling
@jvierling
If you need the importer to support any features that it can't currently handle, then let me know so that I can implement them.
Andreas H. From
@andreasfrom
Adding axioms one at a time and doing a couple of loops in between improves spind a bit. I think testing axioms before adding them might potentially help a lot. (timeout also increased from 30s to 45s)
mode
analytic_independent    17
analytic_sequential     40
spind                   15
treegrammar              7
Stefan Hetzl
@shetzl
@andreasfrom This looks better!
I could very well image that testing is quite useful.
It may also be worthwhile to have a closer look at the difference between independent and sequential induction axioms, maybe spind can somehow emulate the use of the sequential induction axioms
Jannik Vierling
@jvierling
@andreasfrom I have updated the tip tool on the server. The problems should now parse.
Andreas H. From
@andreasfrom
@shetzl Yes, I will look at what @jannikvierling sent me and take a closer look.
@jannikvierling Great, thank you!
Gabriel Ebner
@gebner
@andreasfrom Are you planning on doing any experiments on clogic89 between now and Thursday?
Gabriel Ebner
@gebner
@andreasfrom I'm running some experiments on clogic89 today. Please don't run anything else at the moment.
Gabriel Ebner
@gebner
@andreasfrom Also using clogic89 today.
Ermine516
@Ermine516
@jannikvierling Sorry I was delayed in message you, Vacation :)
I am back now
Is there a gapt meeting today?
Gabriel Ebner
@gebner
No, happy easter!
Ermine516
@Ermine516
If so I can bring it with me and we can discuss
:D, ah yes that is why I was on vacation, right.
Happy easter to you too
see you next week then
Gabriel Ebner
@gebner
Next week is labor day.
Ermine516
@Ermine516
ok, I should look at my calendar then
Andreas H. From
@andreasfrom
@gebner Sorry, didn't see your message. I only ran some stuff yesterday. If you need it today and tomorrow please let me know.
Gabriel Ebner
@gebner
Nah, everything's already done now. Also the deadline got extended by two weeks, so I'm no longer in such a hurry either.
Stefan Hetzl
@shetzl
@Ermine516 The next regular gapt meeting is scheduled for May 15.
Andreas H. From
@andreasfrom

I have a question about importing TIP problems. The file tip2015/list_return_2.smt2 looks like this:

; List monad laws
(declare-datatypes (a)
  ((list (nil) (cons (head a) (tail (list a))))))
(define-fun (par (a) (return ((x a)) (list a) (cons x (_ nil a)))))
(define-fun-rec
  (par (a)
    (++
       ((x (list a)) (y (list a))) (list a)
       (match x
         (case nil y)
         (case (cons z xs) (cons z (++ xs y)))))))
(define-fun-rec
  (par (a b)
    (>>=
       ((x (list a)) (y (=> a (list b)))) (list b)
       (match x
         (case nil (_ nil b))
         (case (cons z xs) (++ (@ y z) (>>= xs y)))))))
(prove
  (par (a)
    (forall ((xs (list a)))
      (= (>>= xs (lambda ((x a)) (return x))) xs))))

and after fixupAndLoad becomes:

∀x0 ∀x1 head(cons(x0, x1)) = x0,
∀x0 ∀x1 tail(cons(x0, x1)) = x1,
∀x return(x) = cons(x, nil),
lam2 = lam,
∀y ++(nil, y) = y,
∀y ∀z ∀xs ++(cons(z, xs), y) = cons(z, ++(xs, y)),
∀y >>=(nil, y) = nil,
∀y ∀z ∀xs >>=(cons(z, xs), y) = ++(apply1(y, z), >>=(xs, y)),
∀y0 ∀y1 nil != cons(y0, y1),
∀x apply1(lam, x) = return(x)
⊢
∀xs >>=(xs, lam2) = xs
So the apply1 must be introduced by the tip tool, but it does not show up in TipProblem.functions which I am using to generate reduction rules for testing.
Jannik Vierling
@jvierling
The symbol apply1 is stored in TipProblem.uninterpretedConsts, because it has no definition.
Andreas H. From
@andreasfrom
Why doesn't ∀x apply1(lam, x) = return(x) constitute a definition?
Jannik Vierling
@jvierling
Actually by definition I meant a definition in terms of define-fun or define-fun-rec or define-funs-rec. Actually the tip tool adds the definition that you gave as an assertion with a :definition keyword, but currently the tip importer does not inspect these keywords.
Andreas H. From
@andreasfrom
Oh right. Okay, thank you.
Jannik Vierling
@jvierling
Next week when I am back i'll try to spend some time to improve the tip importer.
Andreas H. From
@andreasfrom
It's not urgent for my purposes.
Andreas H. From
@andreasfrom

Here's an update:

mode
analytic_independent    43
analytic_sequential     75
spind                   34
treegrammar             16

All the problems solved by the analytic independent mode are affected by the tip importer thing mentioned above which makes my testing of conjectures less efficient or straight up breaks it. Three of them are solved if I just disable testing.

Spind is currently a little slower in total on the problems they have in common:

spind                118,666ms
analytic_independent 93,527ms
Gabriel Ebner
@gebner
@andreasfrom I'm going to use clogic89 over the weekend again.
Andreas H. From
@andreasfrom
@gebner roger.
Andreas H. From
@andreasfrom
@shetzl We talked about going to and from an expansion proof to reduce the proof, but I've got one here (for isaplanner/prop_57.smt2) which has around 2000 steps as an LK proof and going back from expansion to LK is taking more than 20 minutes. So that doesn't seem feasible(?)
Andreas H. From
@andreasfrom
Wait, I realised I could cleanStructuralRules it first and then it roundtrips quickly.