PsiCompund (number_cell, deduction, limit)

A compound distribution is made up of a "severity" distribution and a "frequency" distribution. Assume the following compound distribution, =PsiBeta(3, 2, PsiCompound(A2)), where A2 = PsiPoisson(100). PsiBeta(3,2) is referred to as the "severity" distribution. The severity distribution is the distribution to be added N times. PsiPoisson(100) is referred to as the "frequency" distribution. The frequency distribution determines the size N of the sum (i.e, how many PsiBeta to sum). N can be a constant but can also be computed at each trial by drawing from a discrete distribution.

The number_cell argument passes the number of random trial values to be summed. Number_cell can be an integer, a cell containing an integer, a formula evaluating to an integer, or a cell containing a discrete distribution.

If a fractional value is passed directly or indirectly (by using a formula) or a continuous distribution is passed to this argument, the result will be rounded down to the nearest integer.

Note: If a discrete distribution is passed to the number_cell argument, the frequency distribution must be formulated in such a way that the trial values generated by the distribution must be greater than 1. If not, trial values < 1 will be set equal to 1.

The value passed to the deduction argument is subtracted from every term of the compound sum which results in a shift of the compound distribution by -N * deduction.

If a trial value is larger than a specified limit, then the trial value is reset to the limit.

For a complete example illustrating how to compute a compound distribution, see the Examples: Simulation and Risk Analysis chapter in the Analytic Solver User Guide.