Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Undetected infinite multi-objective reward #253

Open
Chickenpowerrr opened this issue Oct 29, 2024 · 0 comments
Open

Undetected infinite multi-objective reward #253

Chickenpowerrr opened this issue Oct 29, 2024 · 0 comments

Comments

@Chickenpowerrr
Copy link

The property multi(R{"inf"}min=? [ C ]) should obviously be infinite in the model below. However, the linear programming solver incorrectly reports a result of 0.0.

This seems to be the case since the constraint is not violated according to the code since it is not a maximising objective.

mdp

module infinite
    s : [0..1];
    [inf] s=0 -> true;
endmodule

rewards "inf" 
	[inf] true : 1;
endrewards
PRISM
=====

Version: 4.8.1
Date: Tue Oct 29 11:27:13 CET 2024
Hostname: LAPTOP-9DDD1OSE
Memory limits: cudd=1g, java(heap)=1g
Command line: prism infinite.nm -pf 'multi(R{"inf"}min=? [C])' -lp

Parsing model file "infinite.nm"...

Type:        MDP
Modules:     infinite
Variables:   s

1 property:
(1) multi(R{"inf"}min=? [ C ])

---------------------------------------------------------------------

Model checking: multi(R{"inf"}min=? [ C ])

Building model...

Computing reachable states...

Reachability (BFS): 1 iterations in 0.00 seconds (average 0.000000, setup 0.00)

Time for model construction: 0.008 seconds.

Type:        MDP
States:      1 (1 initial)
Transitions: 1
Choices:     1

Transition matrix: 5 nodes (2 terminal), 1 minterms, vars: 1r/1c/1nd
Total time for product construction: 0.0 seconds.

States:      1 (1 initial)
Transitions: 1
Choices:     1

Transition matrix: 5 nodes (2 terminal), 1 minterms, vars: 1r/1c/1nd

Prob0A: 1 iterations in 0.00 seconds (average 0.000000, setup 0.00)

yes = 0, no = 0, maybe = 1

Computing remaining probabilities...
Switching engine since only sparse engine currently supports this computation...
Engine: Sparse

0 Targets:

1 Rewards:
#0: Rmin=?

Building sparse matrix... [n=1, nc=1, nnz=1, k=1] [0.0 KB]
Creating vectors for yes... [0 x 0.0 KB]
Creating vector for maybe... [0.0 KB]
Creating vector for bottom end components... [0.0 KB]
Initial state index: 0
TOTAL: [0.0 KB]

Building LP problem...
Number of LP variables = 2
Adding extra constraints for bounded objectives...
computing max function for rewards ...
Setting objective...
Solving LP problem...
LP problem solution found; result is 0.000000

LP problem solved in 0.00 seconds (setup 0.00, lpsolve 0.00)

Value in the initial state: 0.0

Time for model checking: 0.004 seconds.

Result: 0.0
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant