Opened 10 years ago

Closed 4 years ago

#3076 closed defect (fixed)

Inconsistent behavior of array vs. non-array connect

Reported by: stuchl4n3k@… Owned by: somebody
Priority: high Milestone: 1.16.0
Component: Backend Version: trunk
Keywords: integrator, arrays, simulation Cc:

Description

Description:
I have two simple models (call them A and B) with components.
When I connect the individual components (variables) like: connect(A.foo1, A.foo2);, the model simulation works as expected.

But in model B I have them inside an array and connecting them like: connect(B.foo[1], B.foo[2]); (in a for loop).
The simulation of B fails, giving me:

stdout | OMEditInfo | <p>/tmp/OpenModelica_stuchl4n3k/OMEdit/ArrayBug.IzhNetNeuronTest2 -port=35181 -logFormat=xml -override=startTime=0,stopTime=100,stepSize=0.2,tolerance=1e-6,solver=dassl,outputFormat=mat,variableFilter=.* -dasslJacobian=coloredNumerical -w -lv=LOG_STATS</p>
stdout | warning | <p>DDASSL had repeated error test failures on the last attempted step.</p>
stdout | warning | <p>can't continue. time = 21.185722</p>
stdout | info | <p>model terminate | Integrator failed. | Simulation terminated at time 21.1857</p>
LOG_STATS | info | <p>### STATISTICS ###</p>
stdout | error | <p>Simulation process exited with code 249</p>

Both models work in Dymola.

How to reproduce:
1) See the attached model.
2) Simulate ArrayBug.IzhNetNeuronTest2 time 0-100 (Dassl tolerance 1e-6, 500 intervals, max integration order 5).

System:
I am using Linux Mint Cinnamo 64-bit, OM 1.9.2+dev (r24023).

Attachments (1)

ArrayBug.mo (1.7 KB ) - added by stuchl4n3k@… 10 years ago.
Buggy model package

Download all attachments as: .zip

Change History (2)

by stuchl4n3k@…, 10 years ago

Attachment: ArrayBug.mo added

Buggy model package

comment:1 by Per Östlund, 4 years ago

Component: *unknown*Backend
Milestone: Future1.16.0
Resolution: fixed
Status: newclosed

This seems to have been fixed at some point, both models now simulate successfully.

Note: See TracTickets for help on using tickets.