workin progress
play

workin progress The M in CoPaR From Partition Refjnement to - PowerPoint PPT Presentation

workin progress The M in CoPaR From Partition Refjnement to Minimization Hans-Peter Deifel Oberseminar 09.06.2020 1 / 26 The M in CoPaR From Partition Refjnement to Minimization Hans-Peter Deifel Oberseminar 09.06.2020 1 / 26 workin


  1. workin progress The M in CoPaR From Partition Refjnement to Minimization Hans-Peter Deifel Oberseminar 09.06.2020 1 / 26

  2. The M in CoPaR From Partition Refjnement to Minimization Hans-Peter Deifel Oberseminar 09.06.2020 1 / 26 workin progress

  3. CoPaR Coalgebraic Partition Refjner 2 / 26

  4. Partition Refjnement 3 / 26

  5. Partition Refjnement 3 / 26

  6. Partition Refjnement 3 / 26

  7. Partition Refjnement 3 / 26

  8. Partition Refjnement 3 / 26

  9. Partition Refjnement Difgerent partition refjnement algorithms: Pajge-Tarjan Hopcroft Markov-Chain lumping Color refjnement 4 / 26 ⇒ Used in countless minimization tools

  10. Partition Refjnement Difgerent partition refjnement algorithms: Pajge-Tarjan Hopcroft Markov-Chain lumping Color refjnement 4 / 26 ⇒ Used in countless minimization tools

  11. CoPaR Coalgebraic Partition Refjner 5 / 26

  12. CoM Coalgebraic Minimizer 5 / 26

  13. Coalgebras Labelled Chains Markov Automata Deterministic Systems Transition Set Set Type Functor type functor successor map states 6 / 26 c X FX

  14. Coalgebras Labelled Chains Markov Automata Deterministic Systems Transition type functor successor map states 6 / 26 c X FX Type Functor F : Set → Set 2 × X A FX = P ( A × X ) R ( X ) . . .

  15. Behavioural Equivalence homomorphism Bisimilarity Weighted Equivalence Language Bisimilarity Identifjed by Coalgebra Homomorphism 7 / 26 type functor successor map states c X FX h Fh Y FY d 2 × X A FX = P ( A × X ) R ( X ) . . .

  16. Coalgebraic Partition Refjnement Generic Algorithm Refjnement Interface Partition 8 / 26 F -Coalgebra Functor F for F

  17. Functor Encoding Encoding map Bags 9 / 26 Encoding for Functor F Abstract Type A of labels ♭ X : FX → B ( A × X ) Example: P ( − ) A = 1 ♭ X ( { x 1 , x 2 , x 3 } ) = { ( ∗ , x 1 ) , ( ∗ , x 2 ) , ( ∗ , x 3 ) }

  18. Functor Encoding Encoding map Bags 9 / 26 Encoding for Functor F Abstract Type A of labels ♭ X : FX → B ( A × X ) Example: R ( − ) A = R ♭ X ( f ) = { ( r, x ) | x ∈ X, f ( x ) = r � = 0 }

  19. Functor Encoding Bags Not Natural 9 / 26 Encoding map Encoding for Functor F Abstract Type A of labels ♭ X : FX → B ( A × X ) Let f ∈ R 2 , f (0) = 0 . 5 , f (1) = 0 . 5 ♭ 2 f { (0 . 5 , 0) , (0 . 5 , 1) } R (!) B ( id , !) ♭ 1 / ( ∗ �→ 1 . 0) { (0 . 5 , 1) , (0 . 5 , 1) }

  20. Functor Encoding Encoding map Bags Refjnement Interface 9 / 26 Encoding for Functor F Abstract Type A of labels ♭ X : FX → B ( A × X ) Type W (abstract, could be ints, reals, trees, …) init : F 1 × B A → W update : B A × W → W × F 3 × W

  21. Syntax Input DX q1: {q2: 0.5, q3: 0.5} q2: {q1: 0.5, q3: 0.5} q3: {q3: 1} 10 / 26

  22. Syntax Input DX q1: {q2: 0.5, q3: 0.5} q2: {q1: 0.5, q3: 0.5} q3: {q3: 1} Output Block 0: q1, q2, q3 10 / 26

  23. Syntax Input DX q1: {q2: 0.5, q3: 0.5} q2: {q1: 0.5, q3: 0.5} q3: {q3: 1} Cooler Output DX q1: {q1: 1} 10 / 26

  24. Minimization Two tasks: Computing the (encoding of) the quotient coalgebra Removing unreachable states 11 / 26

  25. Minimization Two tasks: Computing the (encoding of) the quotient coalgebra Removing unreachable states 11 / 26

  26. Quotient Encoding: Goal quotient encoding coalgebra output coalgebra input map 12 / 26 ♭ X c X FX B ( A × X ) q Fq ? Y FY B ( A × Y ) d ♭ Y

  27. Solution: Minimization Interface merge merge and 13 / 26 merge : B ( A ) → B ( A ) B ( A ) fil S · ♭ {} FX B ( A ) Fχ S 1 fil { 1 } · ♭ F 2 B ( A ) {} B ( A ) fil S ( t )( a ) = � x ∈ S t ( a, x )

  28. Minimization Interface: Examples Powerset Monoid-Valued otherwise Polynomial 14 / 26 merge ( ℓ )( ∗ ) = min (1 , ℓ ( ∗ )) � { Σ ℓ } Σ ℓ � = 0 merge ( ℓ ) = {} merge = id

  29. Victory? group monoid-valued ungroup 15 / 26 curry ·B ( swap ) B ( A × q ) ♭ B ( A ) ( Y ) FX B ( A × X ) B ( A × Y ) merge ( Y ) Fq B ( A ) ( Y ) FY B ( A × Y ) ♭ for B -monoid

  30. 16 / 26 Encoding Assumption Nope! We additionally need the following assumption on ♭ : ♭ X FX B ( A × X ) fil { x } Fχ { x } B ( A ) fil { 1 } ♭ 2 F 2 B ( A × 2)

  31. Encoding Assumption: Counterexample An encoding for Powerset But: Ruled out by assumption 17 / 26 A = 1 + 1 ♭ 2 ( { x 1 , . . . , x n } ) = ( { ( inj 1 ∗ , x 1 ) . . . , ( inj 1 ∗ , x n ) } ) ♭ X � =2 ( { x 1 , . . . , x n } ) = ( { ( inj 2 ∗ , x 1 ) . . . , ( inj 2 ∗ , x n ) } ) ⇒ Lawful merge possible!

  32. Encoding Assumption: Counterexample An encoding for Powerset But: Ruled out by assumption 17 / 26 A = 1 + 1 ♭ 2 ( { x 1 , . . . , x n } ) = ( { ( inj 1 ∗ , x 1 ) . . . , ( inj 1 ∗ , x n ) } ) ♭ X � =2 ( { x 1 , . . . , x n } ) = ( { ( inj 2 ∗ , x 1 ) . . . , ( inj 2 ∗ , x n ) } ) ⇒ Lawful merge possible!

  33. Victory! group ungroup 18 / 26 B ( A × q ) ♭ B ( A ) ( Y ) FX B ( A × X ) B ( A × Y ) Fq merge ( Y ) B ( A ) ( Y ) FY B ( A × Y ) ♭

  34. Victory! group ungroup 18 / 26 B ( A × q ) ♭ B ( A ) ( Y ) FX B ( A × X ) B ( A × Y ) Fq merge ( Y ) B ( A ) ( Y ) FY B ( A × Y ) ♭

  35. Minimization Two tasks: Computing the (encoding of) the quotient coalgebra Removing unreachable states 19 / 26

  36. Minimization Two tasks: Computing the (encoding of) the quotient coalgebra Removing unreachable states 19 / 26

  37. Reachability: Problems “Two tasks” all the way down: Pointed Coalgebras Computing the Reachable Subcoalgebra 20 / 26

  38. Reachability: Problems “Two tasks” all the way down: Pointed Coalgebras Computing the Reachable Subcoalgebra 20 / 26

  39. Reachability: Problems “Two tasks” all the way down: Pointed Coalgebras Computing the Reachable Subcoalgebra 20 / 26 ⇒ Simple matter of programming™

  40. Reachability: Problems “Two tasks” all the way down: Pointed Coalgebras Computing the Reachable Subcoalgebra 20 / 26 ⇒ Simple matter of programming™

  41. Computing the Reachable Subcoalgebra In Set: Standard graph search on canonical graph Canonical Graph does not factorize through For coalgebra: 21 / 26

  42. Computing the Reachable Subcoalgebra In Set: Standard graph search on canonical graph 21 / 26 Canonical Graph τ X : FX → PX τ X ( t ) = { x ∈ X | 1 t − → FX does not factorize through F ( X \ { x } ) Fi − → FX } c τ C For coalgebra: C − → FC − → P C

  43. Goal We want: Canonical graph of original coalgebra and canonical graph of encoded coalgebra to be the same! 22 / 26 c τ C FC P C τ ♭ B ( A × C )

  44. Should be true™ Questment: Encoding contains canonical graph? 23 / 26 Direction: τ ( t ) ⊆ τ ( ♭ ( t ))

  45. Questment: Encoding contains canonical graph? 23 / 26 Direction: τ ( t ) ⊆ τ ( ♭ ( t )) ⇒ Should be true™

  46. Encoded Coalgebra a b c d 0.5 0.5 0 Canonical Graph a b d 24 / 26 Counterexample for τ ( t ) ⊇ τ ( ♭ ( t )) Functor: R ( − )

  47. Requirement: Sub-Naturality Encoding must be sub-natural: now works! 25 / 26 ♭ X X FX B ( A × X ) f Ff B ( A × f ) ♭ Y Y FY B ( A × Y )

  48. Requirement: Sub-Naturality Encoding must be sub-natural: 25 / 26 ♭ X X FX B ( A × X ) f Ff B ( A × f ) ♭ Y Y FY B ( A × Y ) ⇒ now works!

  49. Question: Connection? Assumption 1 Assumption 2 26 / 26 ♭ X FX B ( A × X ) fil { x } Fχ { x } B ( A ) fil { 1 } ♭ 2 F 2 B ( A × 2) ♭ X X FX B ( A × X ) f Ff B ( A × f ) ♭ Y Y FY B ( A × Y )

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend