I am still bothered by the difference this is producing. For instance,
if week 1 is 100 and week 2 is 225, then week 1.5 should be 162.5
instead it is shows 159.375.

Hi. I don't think either answer is wrong. The difference is that you are
estimating the midpoint by using a straight line.
If your total sales from the beginning were a straight line, then from your
description above, week 3 would be
225+125 = 350. Week 3.0 is 375.

If you plot your output, you will see that the output is not a straight
line. That's where the small differences are coming from.

1 100
2 225
3 375
4 550
5 750
6 975
7 1225

As a curious side note, we know that the difference between the recurrence
equation and your linear interpolation equation is fixed at say...the
midpoint.
The total sales (s) at an integer value week (x) is given by:

s(x) = 12.5*x*(x+7)

Your midpoint estimate:
( s(x) + s(x+1) )/2

Recurrence midpoint
s(x+.5)

If you subtract the two equations, you get a constant 3.125.
Therefore, a linear interpolation will always be 3.125 higher at the
midpoint between integers (this specific problem).
We can check your data from above...

159.375 + 3.125 = 162.5
--
HTH. :)
Dana DeLouis
"To understand recurrence, one must first understand recurrence."


wrote in message
oups.com...
Thank you for this one as well.
I am still bothered by the difference this is producing. For instance,
if week 1 is 100 and week 2 is 225, then week 1.5 should be 162.5
instead it is shows 159.375.
Why is this again? And really, is it correct if it isn't producing the
correct answer?



Dana DeLouis wrote:
How many cups of coffee total will I have sold them in say 4.2 weeks.

Hi. I believe the shorter recurrence equation for this particular
problem
is:

=12.5*n*(n + 7)

when n=4.2, it returns 588.

I believe it's a little different than 590 because the 200 in .2*200
hasn't
occurred yet.
Hope I said that correctly. :0
--
Dana DeLouis
Windows XP, Office 2003


wrote in message
ups.com...
Nice work, that was it.
If you have a minute, I would love it if you could parse your answer. I
am a pretty basic excel user right now, and it would be neat to get a
better understanding of it's language/formulas.
Thanks again.

JMB wrote:
With your example, this formula gave me 590 for 4.2 weeks

Week1: 100
Week2: 125
Week3: 150
Week4: 175
Week5: 200

Total = (100+125+150+175+(0.2*200) = 590

Where A1 is the number of weeks and B1=25 (each weeks increase)

=SUMPRODUCT(100+((ROW(INDIRECT("1:"&ROUNDDOWN(A1,0 )))-1)*B1))+((ROUNDUP(A1,0)-1)*B1+100)*MOD(A1,1)

Does this give the results that you are looking for?

" wrote:

Can anyone help me solve a sum of series problem in excel?

For instance:
`sum 1 to n: base + (n-1 * delta)

Or more specifically:
Say I have x customers at week 1, say 100, and each week I increase
my
customers by y, say 25.
Week 1 I have 100
Week 2 I have 125

And each week I sell them a cup of coffee.

How many cups of coffee total will I have sold them in say 4.2
weeks.

Week 1 100 customer = 100 cups this week, 100 total
Week 2 125 customer = 125 cups this week, 225 total
Week 3 150 customer = 150 cups this week, 375 total


What I will end up needing is how many cups of coffee from weeks 5
to
7, which I could simply subtract, but could also be made a part of
the
series, which I can't figure out in excel! ;-)

Any help?
Thanks!