Issue with ShiftSample

[modelica-trac-importer]

Reported by eashan.liyana on 24 Jun 2015 15:40 UTC Hi, I have an issue with ShiftSample. You can find an example attached to do this ticket. I have a model Counter which add 1 to an integer I every second. In my second model I use ShiftSample to shift I from Counter.

For that I did two different models:

1- I is store in an intermediate variable **W** and it's shifted

2-I from Counter is directly shifted.

In the first model I don't need to give the value of resolution.

In the second model I have to give the value of resolution, otherwise the simulation fail.

Am I not using shiftSample the right way ?

Thank you in advance for you anwsers,

Eashan

---

Migrated-From: https://trac.modelica.org/Modelica/ticket/1739

[modelica-trac-importer]

Modified by dietmarw on 23 Jul 2015 07:36 UTC

[modelica-trac-importer]

Modified by otter on 18 May 2016 08:23 UTC

[modelica-trac-importer]

Modified by otter on 5 Aug 2016 05:59 UTC

[HansOlsson]

The previously attached model is as follows:

package Issue

  block Counter
    output Integer c(start=0);
  equation
    c = 1 + previous(c);
  end Counter;

  model Top_ok1
    Counter cpt(c(start=7));
    Integer w, x, y, z;

  equation
    w = cpt.c;
    x = subSample(cpt.c, 2);
    y = shiftSample(
        cpt.c,
        1,
        1);
    z = subSample(w, 2);
  end Top_ok1;

  block Top_ok2
    Counter cpt(c(start=7));
    Integer w, x, y, z;
  equation
    w = cpt.c;
    x = subSample(w, 2);
    y = shiftSample(w, 1);
    z = subSample(y, 2);
  end Top_ok2;

  annotation (uses(Modelica_Synchronous(version="0.92"), Modelica(version=
            "3.2.1")));
end Issue;

My conclusions are:

• It was mislabeled as Modelica_Synchronous, but it is just about the synchronous part of Modelica Language.
• I don't see why resolution is needed in Top_ok1. It might be some issue in some version of some tool.
• The two models differ - specifically z is in Top_ok2 computed from y - but in Top_ok1 computed from w. If that line is changed in Top_ok1 I don't see any difference between the results of the models - at least not in current Dymola.

Thus it seems like a modeling issue - unrelated to Modelica_Synchronous. (Actually making graphical models with Modelica_Synchronous might have avoided this.) There might also be a tool-issue, but that should be reported to the corresponding tool-vendor.

Therefore I close this.