[Libreoffice-bugs] [Bug 110993] New: Chart produces incorrect graph w/ some negative points when using large positive values in source cells

[bugzilla-daemon at bugs.documentfoundation.org]

https://bugs.documentfoundation.org/show_bug.cgi?id=110993

            Bug ID: 110993
           Summary: Chart produces incorrect graph w/ some negative points
                    when using large positive values in source cells
           Product: LibreOffice
           Version: 5.3.1.2 release
          Hardware: x86-64 (AMD64)
                OS: Linux (All)
            Status: UNCONFIRMED
          Severity: normal
          Priority: medium
         Component: Calc
          Assignee: libreoffice-bugs at lists.freedesktop.org
          Reporter: h.k.ghost at gmail.com

Created attachment 135050
  --> https://bugs.documentfoundation.org/attachment.cgi?id=135050&action=edit
LibreOffice Calc Chart Plotting Bug

DESCRIPTION

When creating a chart with very large numbers in cells, the line graph produced
is not entirely correct. While the values in the sheet cells are positive and
appear correct, some of them get plotted to the Y-axis incorrectly.

While trying different scaling values for the Y-axis, I saw that some Y-axis
entries were showing up as negative values instead of positive.


STEPS TO REPRODUCE

Create a data set meant to produce a graph to show how quickly common terms in
time complexity grow as the values of N grow.

1. Create a column for values of n, numbered: 1, 10, 20, 30, ... 100
2. Create additional columns with formulas to calculate 2^n and n!, as in:
O(2^n) and O(n!)
2.a. To calculate 2^n, use: =POWER(2, <cell>)
2.b. (OPTIONAL) Add a few smaller columns (e.g. n^2, n log n, etc) for
comparison
3. Select all the data and create a scatter chart to plot all of them
4. Adjust the maximum values for X axis to 100 and Y axis to 1.4E+020
4.a. Double click the table, then the X or Y axis, then the Scale tab.
4.b. Uncheck the "Automatic" checkbox for "Maximum" and set 100


EXPECTED RESULT

The asymptote for 2^n should be one single and continuous, though significantly
steeper relative to n^2, curve towards positive Y. The asymptote for n! should
be the steepest. All values for 2^n, n!, etc. must be plotted on the positive
Y-axis.


ACTUAL RESULT

There're some portions of the curves that drop below into the negative Y-axis,
which is incorrect.

Notice that the cells containing the values used by the chart are all positive,
but the chart does show the graph going through the negative Y-axis.

-- 
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/20170801/6fc60d74/attachment.html>