OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

office message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: [OASIS Issue Tracker] Commented: (OFFICE-2010)chart:class=chart:surface regression



    [ http://tools.oasis-open.org/issues/browse/OFFICE-2010?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16862#action_16862 ] 

Andreas Guelzow  commented on OFFICE-2010:
------------------------------------------

Issue 1:

My reading of ODF 1.1 permits the surface for a surface chart be specified with a single series by giving a rectangular area as teh value of the chart:values-cell-range-address attribute. 

This would allow several surfaces to be specified in the same chart by giving multiple series.
This also matches the description of chart:series in 1.2:
"The <chart:series> element represents a data series in a chart. If the chart requires more input data, like it is the case for scatter and bubble charts, <chart:domain> sub-elements shall be defined that contain the cell range addresses of the corresponding data."
and corresponds to the simlar bubble charts. (Conceptually bubble charts are very similar to contour plots, the former showing the values by the size of the bubbles the latter by the colour used.)

The 1.2 draft appears to change that to require multiple series for the same surface.

There are several issues with this:
1) This prohibits the inclusion of several surfaces in the same chart since it would not be clear how the series are grouped into the surfaces.
2) At least one existing implementation (Gnumeric) saves surface charts using the single-series method.  (Are there any implementations that use the multiple-series method.)

Issue 2: 

In ODF 1.2 chart:three-dimensional unnecessarily changes the interpretation of the axes: 

for chart:three-dimensional="false" the values in the series are the height (shown as contour colours) and relate to the z-axis. (personally I find this to be the natural naming since typically z=f(x,y))

for chart:three-dimensional="true" the values in the series are the height of the surface over the base and relate to the y-axis.
(this corresponds to y=f(x,z) which I find quite unusual).

Note that my main issue here is that chart:three-dimensional changes relationship to the axes, i.e. toggling the value of chart:three-dimensional does not simply changes between contour and surface but also affects the interpretation of axes.

Issue 3: 

In both 1.1 and 1.2 the grid values are fixed to 1,2,3,...,n, these should be specifiable through twp chart:domain elements as for bubble charts. Of course this could be deferred to ODF Next. (Gnumeric currently uses 2 chart:domain elements for this purpose in analogy to the bubble chart type.)

> chart:class=chart:surface regression
> ------------------------------------
>
>                 Key: OFFICE-2010
>                 URL: http://tools.oasis-open.org/issues/browse/OFFICE-2010
>             Project: OASIS Open Document Format for Office Applications (OpenDocument) TC
>          Issue Type: Bug
>          Components: Chart
>    Affects Versions: ODF 1.2
>            Reporter: Andreas Guelzow 
>            Assignee: Patrick Durusau
>             Fix For: ODF 1.2
>
>
> In ODF 1.1 the description for chart:class=chart:surface is:
> the data points are interpreted as tabular data, where each value defines a 'height' at a specific grid location. 
> In the ODF draft 1.2 this has changed to:
> The values of multiple <chart:series> (marked as being of type chart:surface) are interpreted as a 'altitude' at a specific grid location.
> So in 1.1 one needs only one series to give the tabular data (table range) for a whole surface where the column/row location indicates the grid location for this "height".
> In 1.2 on the other hands an undetermined number of series are supposed to be combined in an undetermined way to create the surface. (Some appear to interpret this to mean that each series gives the heights for one fixed coordinate and the series contains rather than tabular data a sequence of heights for the various values of the other coordinates.)
> As a consequence the 1.1 description allows several surface plot to be placed in the same chart. SInce there is no way to determine which series would belong to which surface, in 1.2 this has been removed.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://tools.oasis-open.org/issues/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]