Threaded Mode | Linear Mode

Tim Wessman · (This post was last modified: 05-03-2017 07:41 PM by Tim Wessman.)

I didn't see your question before.

The stat editor is not a spreadsheet. It is a list editor for data. Having live updating columns creates ALL sorts of problems as you can very easily create recursive data (which you actually just did by doing this attempt).

You have a data column of C1/C2 in your symb defining your fit. PredY attempts to evaluates your fit, which evaluates your data that contains PredY, so it then attempts to evaluate your fit which evaluates your data that contains PredY, which then attempts evaluate your fit which evaluates your data that contains PredY, which then attempts to evaluate your fit which evaluates your data that contains PredY.... Smile

No, there generally is NOT need to have live updating *data* in your stat columns. Your data is fixed. It does not change. Modifying the *data* is a really bad thing to be doing when looking at statistics in general.

You can define your DEFINITIONS in symb as those *are* formulas meant to be evaluated. The definitions there are not just C1, C2 and so on, but can be things like LOG(C1) or C1*C2+5 or even define S2 as C1, PredY(C1) or C1, Resid(S1).