Bug Hunting Session
Bug 71799 - CONFIGURATION: New smoothing equation for charting functions (enhancement)
Summary: CONFIGURATION: New smoothing equation for charting functions (enhancement)
Status: NEW
Alias: None
Product: LibreOffice
Classification: Unclassified
Component: Calc (show other bugs)
Version:
(earliest affected)
4.1.3.2 release
Hardware: Other All
: medium enhancement
Assignee: Not Assigned
URL:
Whiteboard: BSA
Keywords:
Depends on:
Blocks: Chart-Enhancements
  Show dependency treegraph
 
Reported: 2013-11-19 17:16 UTC by oregonraters
Modified: 2019-03-02 05:01 UTC (History)
0 users

See Also:
Crash report or crash signature:


Attachments
Two charts demonstrating the problem (33.46 KB, image/jpeg)
2013-11-19 17:16 UTC, oregonraters
Details

Note You need to log in before you can comment on or make changes to this bug.
Description oregonraters 2013-11-19 17:16:13 UTC
Created attachment 89480 [details]
Two charts demonstrating the problem

When charting functions, a cubic spline curve gives counter-intuitive results. In the attached, the first curve would make a reasonable person wonder, who are the outliers? (There are none, all zeros) In the second curve, there seems to be a lot of excess wobbling - you would expect Jan-Mar to be straight, and the line to not go higher than April, but the line peaks /before/ April.

This seems to be caused by the Cubic Spline function treating the X and Y axises as equal. It would be nice if there was a new "functional interpolation" line, smoothing that was geared more for functions.

This is discussed a little here (http://ask.libreoffice.org/en/question/24872/custom-cubic-spline-smoothing-equation-for-line-in-chart/).

Steps to reproduce:
1. Chart a smooth-line, line chart of the points 0,0,0,0,0,20000,0 (see attachment)

Current behavior:
wobbles a lot, apogee of curve is AFTER the 20000 point, the line goes (a little) higher than the highest point, 20000.


From Regina on (http://ask.libreoffice.org/en/question/24872/custom-cubic-spline-smoothing-equation-for-line-in-chart/) "
You are likely looking for a functional interpolation, where the points, which are added by the interpolating algorithm, behave the same as if they were true functions points, and where y depends on x. Such interpolation does not exist in LO. But I agree, that such kind of interpolation is needed. There already exists suitable code in OOo1.1.5 using splines, but other ways are perhaps better. If LO has an implementation it will be possible to get it later on into ODF standard.

              
Operating System: All
Version: 4.1.3.2 release
Comment 1 Owen Genat (retired) 2013-11-19 22:11:00 UTC
I have commented in the original AskLO thread, which contains plenty of detail, although the comment from Regina (quoted above) is the key piece in terms of this enhancement. Confirmed. Status set to NEW. Severity set to enhancement. Maybe version should be set to Inherited from OOo?
Comment 2 oregonraters 2013-11-20 01:23:26 UTC
Set to "Inherited from OOo".
Comment 3 Owen Genat (retired) 2014-01-06 09:59:13 UTC
Setting version back from "Inherited From OOo" to what it was prior (for bugs marked enhancement), for QA tracking purposes. My mistake. Apologies.
Comment 4 7qia0tp02 2019-03-02 05:01:18 UTC
It would be nice if the smooth curve used https://en.wikipedia.org/wiki/Monotone_cubic_interpolation so it didn't overshoot between points.