extend.ranvar

extend.ranvar(R : ranvar) 🡒 U.*, table function

Extends a ranvar into a table containing the inclusive segments of its buckets.

Example:

x = poisson(3)
table G = extend.ranvar(x)
show table "poisson(3)" a1c5 with
  G.Min
  G.Max
  int(x, G.Min, G.Max) as "Probability"

The vectors Min and Max are present in the returned table which extends the table passed as argument. Those two values represent respectively the lower and higher inclusive boundaries of each segment. The segments are contiguous, and always include [0, 0].

extend.ranvar(R : ranvar, Gap : number) 🡒 U.*, table function

This overload of extend.ranvar (see above) ensures that both segments [0, 0] and [1, gap] are found in the resulting list of segments. Later segments are generated starting from gap + 1.

Example:

x = poisson(3)
gap = 4
table G = extend.ranvar(x, gap)
show table "poisson(3)" a1c5 with
  G.Min
  G.Max
  int(x, G.Min, G.Max) as "Probability"

This overload is typically intended to be used with the sum of the stock on hand plus the stock on order as the gap.

extend.ranvar(R : ranvar, Gap : number, Multiplier : number) 🡒 U.*, table function

This overload of extend.ranvar (see above) ensures that

Example:

x = poisson(3)
gap = 4
multiplier = 2
table G = extend.ranvar(x, gap, multiplier)
show table "poisson(3)" a1c5 with
  G.Min
  G.Max
  int(x, G.Min, G.Max) as "Probability"

This overload is typically intended to be used with lot multiplier which are sometimes found among purchasing constraints.

extend.ranvar(R : ranvar, Gap : number, Multiplier : number, Reach : number) 🡒 U.*, table function

This overload of extend.ranvar (see above) ensures that

Example:

x = poisson(3)
gap = 4
multiplier = 5
reach = 20
table G = extend.ranvar(x, gap, multiplier, reach)
show table "poisson(3)" a1c5 with
  G.Min
  G.Max
  int(x, G.Min, G.Max) as "Probability"

This overload is typically intended to be used to cope with MOQs (Minimum Order Quantities). Indeed, as the MOQ may exceed the support of the ranvar, the segments may not go far enough to reflect any eligible purchasing decision. This reach overload mitigates this issue.

As a rule of thumb, we suggest not to use this overload unless there are specific MOQs to be reached. When this overload has to be used, it is suggested to keep reach as small as possible to limit the compute overhead. A small reach value does not prevent higher segments to be produced.

See also