The set of dialog items you find in the 窶廢vent窶 view varies from application to application. However, two items stay unchanged: the 窶廰.Fact.窶 field, and 窶廚ondition窶 popup menu.
Unlike nodes, an event does not necessarily have to be calculated in every simulation run. DynRisk allows you to attach a condition to each event. The condition is defined by the 窶廰.Fact.窶 field and the 窶廚ondition窶 popup menu. When the condition is not satisfied, no message is sent to the target application, and the output value of the event is zero.
The following condition types are available:
窶｢ On all runs
窶｢ On large input
窶｢ On small input
窶｢ On random
窶｢ On even runs
窶｢ On odd runs
窶｢ On init
The 窶廾n all runs窶 option implies that the event is calculated in every simulation run. In this case the 窶廰.Fact.窶 is not used at all.
窶廾n large input窶 means that the event will only be calculated if the sum of the input values is strictly greater than the value of the 窶廰.Fact.窶 field.
窶廾n small input窶 means that the event will only be calculated if the sum of the input values is strictly smaller than the value of the 窶廰.Fact.窶 field.
If the 窶廾n random窶 option is selected, then the 窶廰.Fact窶 field is interpreted as a probability. Thus, when using this option, you should specify a number between 0 and 1 in this field. In each simulation run the event will be calculated with probability equal to the number you specify.
窶廾n even runs窶 means that the event is only calculated every second simulation run starting at run 2. Thus, the event will be calculated in run 2, 4, 6, 窶ｦ, and not calculated in run 1, 3, 5, 窶ｦ. In this case the 窶廰.Fact.窶 field is not used at all.
窶廾n odd runs窶 means that the event is only calculated every second simulation run starting at run 1. Thus, the event will be calculated in run 1, 3, 5, 窶ｦ, and not calculated in run 2, 4, 6, 窶ｦ. In this case the 窶廰.Fact.窶 field is not used at all.
If the 窶廾n init窶 option is selected, the event is only calculated during initialization. The initialization is carried out before any actual simulation is done. At this early stage there is no data flow through the model. Thus, in particular, there are no input values coming in to the event from its predecessors, and no output values being passed on to its successors. To understand what the event does, we need to know what category the event belongs to:
If the event is a 窶彜end data窶 event, then it sends its 窶廰.Fact.窶 value. During the actual simulations, the event will pass zero on to each of its successors.
If the event is a 窶彝eceive data窶 event, it stores the received value. During the actual simulations, the event will pass this value on to each of its successors (instead of just zero). In this case the 窶廰.Fact.窶 field is not used at all.
If the event is a 窶彜end and receive data窶 event, it sends its 窶廰.Fact.窶 value, and stores the received value. During the actual simulations the event will pass this value on to each of its successors (instead of just zero).
If the event is a 窶弋rigger action窶 event, the event simply send the trigger message. During the actual simulations the event will pass zero on to each of its successors. In this case the 窶廰.Fact.窶 field is not used at all.
