
Re: Interpolation of 2D AF Tables
Bannikov Nov 21, 2013 1:01 PM (in response to vincent.spaa)Hello Vincent,
Currently, as far as I know, there isn't possible to do this without some custom development. Some time ago we've developed a custom AF Data Reference (as I remember, even for AF 1.x) to do similar task. You can follow this way or wait for OSIsoft to implement it in furure version of AF (1D interpolation is already on its way to release).

Re: Interpolation of 2D AF Tables
vincent.spaa Nov 21, 2013 1:05 PM (in response to Bannikov)Hello Sergey,
Thanks for your reply. Is that custom AF DR still around somewhere? Or is it not compatible with the current version of AF?

Re: Interpolation of 2D AF Tables
skwan Nov 22, 2013 5:20 AM (in response to vincent.spaa)Vincent:
Is your need restricted to linear interpolation of X and Y or do you have a need for a curve fit? (e.g. curve fit the Y values using a quadratic or even higher order polynomials)

Re: Interpolation of 2D AF Tables
vincent.spaa Nov 22, 2013 10:30 AM (in response to skwan)Hello Steve,
Thank you for response.
Linear interpolation should suffice for our usecase. Is AF capable of doing that outofthebox? Or is there a custom AF DR available?

Re: Interpolation of 2D AF Tables
cmanhard Nov 25, 2013 1:20 PM (in response to vincent.spaa)AF does not have a 2D Interpolation function in the Table Lookup DR, nor am I aware of a custom DR. You can create one using the existing Table Lookup combined with Formula, but it is not pretty. It takes 8 table lookups and 3 formulas. With the upcoming Abacus, scheduling the final calculation as a PE would ensure the data is uptodate with less impact on the clients as they would not need to reexecute the 8 table lookups.
Table has columns, x, y, and R.
Attributes X and Y are the inputs, Attribute R is the Result
X.Min and X.Max are the bounding values of input X in the table's x column
Y.Min and Y.Max are the bounding values of input Y in the table y column
R1R4 are the 4corner results in the table which bound the result
RA and RB are used in the final calculation of R



