[Libreoffice-bugs] [Bug 127517] Libreoffice Calc Trendline R^2 value calculation goes wrong when trendline is forced through 0

[bugzilla-daemon at bugs.documentfoundation.org]

https://bugs.documentfoundation.org/show_bug.cgi?id=127517

Buovjaga <ilmari.lauhakangas at libreoffice.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jumbo4444 at yahoo.fr

--- Comment #6 from Buovjaga <ilmari.lauhakangas at libreoffice.org> ---
(In reply to Leyan from comment #5)
> I had a look in the code, this is deliberate. A specific computation was
> introduced to reproduce the behavior of the LINEST function, by replacing
> the difference to the average by the difference to the constant to intercept:
> 
> * The original bug is here:
> https://bugs.documentfoundation.org/show_bug.cgi?id=73374
> * The commit made to solve it is here:
> https://github.com/LibreOffice/core/commit/
> 9e67015be6e9e53e19466117d692de17c18463da
> 
> The LINEST function also gives R2=0.985148.
> 
> In Excel (Office365), the LINEST function answers 0.985148 and the chart R2
> is 0.9373.

Laurent: can you comment on this?

-- 
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/20200830/01974325/attachment.htm>