Re: calculate plot points for a skewed distribution curve / histogram


obiron wrote:
Hi guys, I'm hoping you can help me here.

I am trying to build a function that will allow me to plot the points
on a curve for a skewed distribution.

The curve is to predict sales volumes for a marketing campaign

I know:-

the length of the campaign in days (x-axis)
The total number of sales expected (area under the curve)
Therefore I can calculate the mean...
The largest number of sales in a given day (Modal average)
The day that this number of sales is expected to occur
The upper an lower bounds for 90% confidence factor.

as an example:-

Campaign length = 180 days
Total sales expected = 1200 units

Mean average is therefore 6.67 units per day
Maximum sales will be 30 units and this will occur on day 22
The 90% confidence factors lower bound is 7 days and the upper bound is
45 days.

So I know the graph is going to be skewed left with a very long tail.

How do I calculate the predicted sales for each day?

A beta-distribution with alpha=2.6 and beta=12.5 has its
mode at 22/180 with height 30/(1200/180) but doesn't
really have a long tail.

.

Relevant Pages

  • calculate plot points for a skewed distribution curve / histogram
    ... on a curve for a skewed distribution. ... The curve is to predict sales volumes for a marketing campaign ... The upper an lower bounds for 90% confidence factor. ...
    (sci.stat.math)
  • Re: DABSWTFM - lost his touch?
    ... "if you'd graphed the sales themselves instead of the sales ... growth if would be an inverted hockey stick curve" ... standard curve, I can't think of the name, starts off shallow, picks ... graph, ...
    (alt.radio.digital)
  • Re: DABSWTFM - lost his touch?
    ... "if you'd graphed the sales themselves instead of the sales ... growth if would be an inverted hockey stick curve" ... standard curve, I can't think of the name, starts off shallow, picks ... Theres more to algorithms than just maths. ...
    (alt.radio.digital)