Dear Lori,

That's very interesting. I'll confess I actually haven't looked at
this in quite a while.

I just read what Wiki says about Catmull Splines. It looks like that
uses central difference to get the slopes. I don't remember what bits
of logic I had implemented for my spline.

I tried to use your IRR cell formula but only got #N/A out of it. So
I couldn't compare it to what my function returns. If you have a
workbook that shows a comparison, you can send it to me if you wish.
What "issues" with my function does your solution resolve?

Brian





On Apr 28, 1:28 am, "Lori Miller"
wrote:
Brian - off topic, i saw a thread from you a while ago about how
Excel calculates smooth curves and created a few formulas and
simplified code, with good results. See links in this thread:http://help.lockergnome.com/office/d...ct1005590.html

The values are the same as yours in general but the correction
factor differs slightly. I'm pretty sure this adjustment is exact.and
has a natural interpretation in terms of tangent conditions. I'm
also fairly certain Excel uses de Casteljau's algorithm to calculate
the curve with straight lines joining the calculated points.

"Brian Murphy" wrote in message

...

In VBA I can create a user defined Type that contains fixed size
arrays, and pass one of these to a C DLL where it goes into a struct,
if the struct was compiled with the correct array sizes ahead of time.

Is it possible to do this with variable size arrays?

Thanks,

Brian Murphy