Microsoft Office Tutorials and References

In Depth Information

Exhibit 8.19
Two approaches to arrivals

approaches for simulating the arrivals in a 2 hour period. The range (A1:J25) has

two separate VLOOKUP functions sampling a Poisson with an average arrival rate

of one unit per hour. Why two VLOOKUPs? We have two because one is needed for

each hour in the two hour period. This is the approach we used in Exhibit 8.18. The

approach in range K1:T25 is a single sample which is then multiplied by two. Why

a single sample? In this case we assume that the same value can be used for each

of the two hours. Both approaches collect 250 samples of arrivals for a two hour

period.

The summary statistics in Y27:Z30 show very similar behavior in terms of means

and total arrivals for the 250 samples, approximately a mean of 2.0 and total arrivals

500 for both. Thus, both approaches appear to be similar in some summary statistics.

The difference is seen in the standard deviation for the two VLOOKUP approaches.

The approach with two randomly sample hours, has a lower value, 1.4, than the

approach which used a single VLOOKUP multiplied by two, 2.1. The graph in

Exhibit 8.19 shows the greater dispersion of results for the approach that gener-

ates a single VLOOKUP sample then multiplies by 2. It is far more likely that this

Search JabSto ::

Custom Search