We'll let the OP decide what's best...

I have little to add other than refer to Harlan's comments on this:
http://groups.google.com/group/micro...bc40ca23ed8331

sigmoid function. If you want to use a smooth function, try a
cubic spline then R²=1 exactly and you can set SHGC=0 at A=90
as in the diagram giving better behavior at the boundary.

"Bernd P" wrote in message
...
Hello Lori,

I honestly think that if your approach is published without any
information about its restrictions users are being lead up the garden
path.

I took the liberty to publish this as a bad example:
http://www.sulprobil.com/html/interpolate.html

Regards,
Bernd