Load attached file; select the data range, click 'insert chart'.
Notice the horrible Y scale that is selected - most odd :-) is there some X/Y confusion going on ?
Created attachment 55225 [details]
and confirming :-)
[Reproducible] with Parallel Dev-Installation of "LibreOffice 3.5.0 Beta2- WIN7 Home Premium (64bit) German UI [Build-ID : 8589e48-760cc4d-f39cf3d-1b2857e-60db978] and reporter's sample, Y-max is "45000" although maximum value is only "5177"
Has already been submitted as "Bug 44437 - FILEOPEN particular .xls shows wrong scale for y-axis", but this report seems to show that it's a more general problem.
It seems that the problem has to do with the x-axis values, the maximum y-axis value would be appropriate for the maximum x-value (format as standard number); but there seem to me additional conditions for the problem.
Creating a new LINE chart in sample document with 3.4.5 first a lines chart will be shown with y-max 45000 and an x value line above 40000, when you click "first column as label" X-values line disappears and view switches to Y-axis max=6000
Creating a new LINE chart in sample document with 3.5.0 it starts as in 3.4.5, when you click "first column as label" X-values line disappears as in 3.4.5, but Y-axis max remains.
With what did the problem start?
[Reproducible] with Server installation of Master "LOdev 3.5.0beta2+ – WIN7 Home Premium (64bit) ENGLISH UI [Build ID: 22905fb-b605e4f-4c1bcb5]" Win-x86@6- - pull time 2011-12-20 20:59:59)
Here somewhere the problem started Between Beta1 and Beta2?
Worked fine with parallel installation of MinGW Master "LibO-dev 3.5.0 Beta 1 – WIN7 Home Premium (64bit) English UI (Build ID: 8935521-b204871-3e50423-4c1bcb5) (daily/Win-x86@7-MinGW/master/2011-12-13_03.17.52)"
Worked fine with Server installation of MSVC Master "LibO-dev 3.5.0 – WIN7 Home Premium (64bit) ENGLISH UI [Build ID: a286353-090bcba-3bf3b94]" Win-x86@6 – 2011-12-02_22:36:35)
Worked fine with Server installation of Master "LibO-dev 3.5.0 – WIN7 Home Premium (64bit) English UI [(Build ID: 5d1a991-4cb1bac-ca7e6f5-9125509-ce71330)]" (111109)
Please feel free to reassign (or reset Assignee to default) if it’s not your area or if provided information is not sufficient. Please set Status to ASSIGNED if you accept this Bug.
*** Bug 44437 has been marked as a duplicate of this bug. ***
Attachment from Bug 44430 which Bug 44437 refers to is located here (just to make it easier to get it without jumping to two other bugs to get to it)
I'll look into it.
Ha ha! I already know even without looking at the code what we are doing wrong. We are adding the y values whose x value is identical, for scale estimation. I added that change to fix the stacked chart scale estimation bug.
The correct logic was somewhat more complex than I thought, but hopefully I got it (mostly) correctly. There may still be some corner cases where it is not optimal, but it's still better than before.
The fix will go into 3.5.0 Beta3.