detect completely constant parts

[Willi Braun]

We evaluate completely constant parts again an again, this should be avoided by detecting part that are independent on time, states or events. In particular we should avoid that for bigger non-linear loops. For example the non-linear loop below will solved 10000!

model M
  Real x,y;
equation
  x = sin(y);
  y = cos(x);
 annotation (experiment(StopTime=10, NumberOfIntervals=10000));
end M;

[Jens Frenkel]

This should not be too complicated. The systems are seperated by partition independent blocks. If there is a equation system with no state or event it could be solved before starting the integration. Maybe it could also solved during compilation time but therefore code generation or an interpreter is needed.

[Willi Braun]

No, it's not that complicated, but just to check for states and events is not enough.
You need too check the dependencies for all variables that are involved in the system
If all of them are independent of time and states the system should be constant.

[Martin Sjölund]

Postponed until 1.9.1

[Martin Sjölund]

This ticket was not closed for 1.9.1, which has now been released. It was batch modified for milestone 1.9.2 (but maybe an empty milestone was more appropriate; feel free to change it).

[Martin Sjölund]

Milestone changed to 1.9.3 since 1.9.2 was released.

[Martin Sjölund]

Moved to new milestone 1.9.4

[Martin Sjölund]

Milestone pushed to 1.9.5

[Martin Sjölund]

Milestone renamed

[Martin Sjölund]

Ticket retargeted after milestone closed

[Martin Sjölund]

Milestone moved to 1.12.0 due to 1.11.0 already being released.

[Francesco Casella]

I tried this with v1.13.0-dev-155-g68350e9

model M
  Real x,y;

  function f_sin
    input Real x;
    output Real y;
  algorithm
    y:=sin(x);
    Modelica.Utilities.Streams.print("computed f_sin");
  //annotation(Inline = false);
  end f_sin;
  
equation
  x = f_sin(y);
  y = cos(x);
 annotation (experiment(StopTime=10, NumberOfIntervals=100));
end M;

and in fact the string is still written to the output 100 times during simulation.

The detection of the time-invariant variables should be improved.