We have been shortlisted! We are delighted to be nominated for 5 Tech for Good awards for our work with South African charity Shout It Now. Read more here.
Re-sampling Overruns Between Time Slots
Click this check box in the Time-Dependent Distributions Time-Slot dialog if the value being sampled represents a time and if you want times sampled in the previous cell not to overlap into the current cell.
The default is OFF (not checked). Only use this facility if you are sure you need it because it can distort the results collected from your simulation if used when not required. See the example below.
In this example of arrivals to a hospital's “accident and emergency” unit the time interval between arrivals is very different at different times of day. Rush hour a is very busy time.
|Time Slot (start)||Average Interval Between Arrivals (minutes)|
|17:30||5 (Should Resample Overruns)|
Without Re-Sampling Overruns set on the 17:30 - 18:30 period a sample taken at time 17:29 could easily result in the arrivals jumping right over the 17:30 - 18:30 period.
Don't use this facility if there is only a small difference between sample values in adjacent slots because to do so tends to reduce the average sample value (large numbers tend to get rejected unnecessarily).
Don't use this facility if the value sampled is not used for a time or if a sampled time will not affect when the next sample is taken. (It usually only applies to distributions used for inter-arrival times (in Work Input Objects).