Issue 12561: Section: Annex D.2 (marte-ftf) Source: (, ) Nature: Enhancement Severity: Significant Summary: We have found that the current format for modelling probability distributions is not very good is some cases. For example, it is hard to model correctly the probability distribution related to a SporadicPattern. We propose that distributions are modelled as a ProbabilityDistribution <<choice_type>> element, with each probability distribution modelled as a <<tuple_type>> element each with the parameters needed (e.g. Uniform <<tuple_type>> should have 2 parameters: a min:NFP_Real and max:NFP_Real). Resolution: There is a misunderstanding in the way MARTE deals with probability distribution expressions. This is, in any case, a weak aspect in the description of such specification mechanism in MARTE. This resolution proposes to clarify this aspect, by adding an example in the NFP chapter and by making explicit the list of distributions in Annex D. In addition, this issue proposes to complete the textual description of all the NfpTypes included in the library MARTE_NfpTypes. Issue Dependency Warning: Note that the texts and figures provided in this issue depend on Issue 12196, which proposes new probability distributions. Thus, if Issue 12196 is not accepted, these new probability distributions need to be removed from the texts and figures proposed in this issue. Revised Text: see ptc/2009-05-12 pages 217 - 222 Actions taken: July 3, 2008: received issue October 16, 2009: closed issue Discussion: End of Annotations:===== m: webmaster@omg.org Date: 03 Jul 2008 05:52:26 -0400 To: Subject: Issue/Bug Report -------------------------------------------------------------------------------- Name: Adrián Noguero Company: ESI mailFrom: adrian.noguero@esi.es Notification: Yes Specification: A UML profile for MARTE Section: Annex D.2 FormalNumber: ptc/07-08-04 Version: Beta 1 RevisionDate: 04-08-2007 Page: 434-438 Nature: Enhancement Severity: Significant HTTP User Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; ESI Internet Explorer; .NET CLR 1.1.4322; InfoPath.1) Description We have found that the current format for modelling probability distributions is not very good is some cases. For example, it is hard to model correctly the probability distribution related to a SporadicPattern. We propose that distributions are modelled as a ProbabilityDistribution <> element, with each probability distribution modelled as a <> element each with the parameters needed (e.g. Uniform <> should have 2 parameters: a min:NFP_Real and max:NFP_Real).