[Libreoffice-bugs] [Bug 121396] FORECAST.ETS.PI functions produces different results with identical data.

[bugzilla-daemon at bugs.documentfoundation.org]

https://bugs.documentfoundation.org/show_bug.cgi?id=121396

--- Comment #9 from Winfried Donkers <winfrieddonkers at libreoffice.org> ---
Created attachment 146625
  --> https://bugs.documentfoundation.org/attachment.cgi?id=146625&action=edit
document used for testing

I have not yet been able to reproduce a crash with the attached document,
neither with direct (re)calculation (single cells and groups of cells) nor with
the function wizard. The attached document is an extended version of the
original attachment.
I tried 2 computers, 1 with current master, 1 with version 6.0.5.2. Both
computers have openSUSE Leap 15.
I will try tomorrow on a Windows computer.

A note on the document in the attachment: with a data set of 5 samples and a
prediction that is 6 steps away from the latest sample (daily values from 10 to
14 May and a prediction for 20 May) the results are way out of reality from a
statistical point of view. Also, the data set does not contain at least 2
complete periods, which is needed for ETS. This should not influence the
stability of the calculations, it's just that the results are useless.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/libreoffice-bugs/attachments/20181114/9ed20d28/attachment.html>