Hi,
running FDE (on standardized data - though I do not believe this is critical) and creating a B-Spline Model you get Eigenfunctions and FPC, like the below:
Rerunning the Analysis either by save script and run or using rerun option in the red hotspot changes number of Eigenfunctions found as well the FPCs. Though the statistics seem to stay the same:
I haven't seen any random seed setting I could use, if there is used some anywhere in the calculations. I guess it's due to a mathematical non-unique solution. However this is confusing when customers experience this.
Strange is as well that a redo analysis removes all set knots to default (also from the then saved script). I guess this is a bug.
Any insight is welcome.
@clay_barker @chris_gotwalt1
