Hi,

We have a requirement of 3 calculations in AF Analysis - one for data quality checks and 2 others consuming the first one.We are looking at following options to set them up in a better way -

1) Have data quality as a separate calculation that will be used in the other 2 consumer calculations.

2) Merge the data quality checks as part of the equations for the 2 consumer calcs so that the no of analyses can be brought down to 2 instead of 3. But doing this will increase the number of variables in the 2 consumer calcs and might also increase the computational power required to perform 1 calculation (I am not sure whether increasing variables really degrades the performance of the calculation).

Which of the 2 might be better ? I believe, if adding new variables will not degrade the performance, then option 2 might be better.Any ideas ?

Thanks,

Amit

Hi Amit,

Good rule of thumb, an analysis should be configured to perform a single calculation. If it helps to break a complex calculation down into smaller calculation blocks (and maybe output some of the intermediate results) then using variables is the way to go. I wouldn't put two completely independent calculations within the same analysis as this would unnecessarily increase the load on the PI Analysis Service to evaluate this analysis.

Thanks,

Seb