[Scilab-users] xs2svg and complex grayplot: room for improvements?

[amonmayr at laas.fr]

Le 22/08/2018 à 18:53, Samuel Gougeon a écrit :
> Hello Antoine,
>
> Le 22/08/2018 à 15:19, antoine.monmayrant at laas.fr a écrit :
>>
>> Hi all,
>>
>> xs2svg is not working well with big grayplots: the resulting svg file 
>> is too big and contains too many elements to be usable. Worse, for 
>> really big grayplots, the export fails (not enought memory).
>>
>> As an example, the following code:
>>     grayplot()
>>     h=gcf();
>>     xs2svg(h,'grayplot.svg');
>> results in a 3.7MB svg file with 93784 elements that cannot be edited 
>> with Inkscape.
>>
>> The issue is that xs2svg renders each patch of the grayplot using two 
>> svg filled triangles.
>> In "he who must not be named"¹, the svg export is using a more 
>> pragmatic approach: everything is rendered as svg path/text/whatever, 
>> except from the intensity maps (grayplots, surf, ...) that are 
>> rendered as bmp and included in the svg file.
>> This way, the svg file is really light, can be tweaked with any svg 
>> compliant software (inkscape) and the result is really good.
>> In practice, to get publication-quality grayplots, I tend to do the 
>> following:
>> -hide the grayplot, keep all the rest, export to svg,
>> -hide everything but the grayplot, export to png,
>> -use inkscape to include the png of the grayplot and hand-place it in 
>> the svg.
>>
>
> Do you remember Calixte's answer to you about this topic in 2012?:
> http://bugzilla.scilab.org/show_bug.cgi?id=11195#c1

Her, to be honest, no, I dit not remember.
But Matplot() and Matplot1() are not decent solutions, just workarounds 
as they are way more limited than surf(), Sgrayplot(), contourf()...
In particular, you cannot specify x-axis and y-axis scales using 
vectors, which is essential for experimental dataset that are not always 
regularly spaced in x- and y-.
Thanks for refreshing my memory.
But it means the solution exists: can't we use the same approach for 
other "intensity maps"?

Antoine

>
> Cheers
> Samuel
>
>
>
> _______________________________________________
> users mailing list
> users at lists.scilab.org
> http://lists.scilab.org/mailman/listinfo/users


-- 
+++++++++++++++++++++++++++++++++++++++++++++++++++++++

  Antoine Monmayrant LAAS - CNRS
  7 avenue du Colonel Roche
  BP 54200
  31031 TOULOUSE Cedex 4
  FRANCE

  Tel:+33 5 61 33 64 59
  
  email : antoine.monmayrant at laas.fr
  permanent email : antoine.monmayrant at polytechnique.org

+++++++++++++++++++++++++++++++++++++++++++++++++++++++

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.scilab.org/pipermail/users/attachments/20180823/c70135c7/attachment.htm>