<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Mon, Dec 15, 2014 at 11:15 PM, David Brown <span dir="ltr"><<a href="mailto:dbrown@ucar.edu" target="_blank">dbrown@ucar.edu</a>></span> wrote:<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">Hi Mathieu,<br>
I am interested in your comment that "the speed is still not<br>
absolutely excellent at 2.8s". What kind of performance do you think<br>
would be "excellent"? Are there other tools that achieve the<br>
performance you want? We are interested in any actual comparisons you<br>
want to give us. And of course we are always seeking to improve the<br>
performance of NCL as long as it does not compromise accuracy, which<br>
is our over-arching goal.<br></blockquote><div>Ok, I understand your concern. As I said, the 2s is acceptable though not</div><div>absolutely excellent.</div><div><br></div><div>I will write some test for the plot of unstructured grid and compare with other</div><div>programs. I trully want to have unstructured mesh plots and the situation</div><div>is not excellent right now.</div><div><br></div><div>However, I can already give you a test where ncl is problematic.</div><div>I want to draw a finite element mesh on the plane. So basically 80000</div><div>edges.</div><div><br></div><div>I use "gsn_add_polyline" for each edge which I know is problematic for</div><div>ncl, but I do not see any obvious alternative in ncl for the task. However,</div><div>a same call to SVG (Scalable Vector Graphics) works correctly with</div><div>"inkscape" with similar syntax.</div><div><br></div><div>Test files are available via</div><div><a href="http://mathieudutour.altervista.org/Sending/PLOT_MESH.tar.gz">http://mathieudutour.altervista.org/Sending/PLOT_MESH.tar.gz</a></div><div><br></div><div> Mathieu</div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
-dave<br>
<br>
On Thu, Dec 11, 2014 at 7:21 AM, Mathieu Dutour<br>
<div class=""><div class="h5"><<a href="mailto:mathieu.dutour@gmail.com">mathieu.dutour@gmail.com</a>> wrote:<br>
> Dear Mary,<br>
><br>
> thank you very much! Yes with "RasterFill" it works correctly.<br>
> I used version 6.2.1.<br>
><br>
> However the speed is still not absolutely excellent at 2.8s for a<br>
> plot. Is there a list of standard practice to make plots faster?<br>
><br>
> Mathieu<br>
><br>
> On Thu, Dec 11, 2014 at 7:13 AM, Mary Haley <<a href="mailto:haley@ucar.edu">haley@ucar.edu</a>> wrote:<br>
>><br>
>> Mathieu,<br>
>><br>
>> What version of NCL are you running?<br>
>><br>
>> ncl -V<br>
>><br>
>> I was able to run your script just fine, but I changed cnFillMode from<br>
>> "AreaFill" to "RasterFill". See attached png.<br>
>><br>
>> --Mary<br>
>><br>
>><br>
>> On Mon, Dec 8, 2014 at 8:17 AM, Mathieu Dutour <<a href="mailto:mathieu.dutour@gmail.com">mathieu.dutour@gmail.com</a>><br>
>> wrote:<br>
>>><br>
>>> Dear ncl users,<br>
>>><br>
>>> I have some errors for some contour plots on unstructured<br>
>>> grids (the script is in attachment and the relevant netcdf file<br>
>>> is available from<br>
>>> <a href="http://mathieudutour.altervista.org/Sending/test.nc" target="_blank">http://mathieudutour.altervista.org/Sending/test.nc</a> )<br>
>>><br>
>>> Below is exact error:<br>
>>> fatal:ContourPlotPreDraw: ARSCAM/ARPRAM - ALGORITHM FAILURE<br>
>>> fatal:ContourPlotDraw: draw error<br>
>>> fatal:PlotManagerPreDraw: error in plot pre-draw<br>
>>> fatal:_NhlPlotManagerDraw: PreDraw error<br>
>>><br>
>>> I believe that the data is too irregular for the algorithm to plot it.<br>
>>> But is there any way to alleviate this problem?<br>
>>><br>
>>> Secondly, the plots sometime takes extremely long time to be done.<br>
>>> The mesh has only 110804 nodes which is nowhere near big. What<br>
>>> would be possible to avoid that problem?<br>
>>><br>
>>> Mathieu<br>
>>><br>
>>> _______________________________________________<br>
>>> ncl-talk mailing list<br>
>>> List instructions, subscriber options, unsubscribe:<br>
>>> <a href="http://mailman.ucar.edu/mailman/listinfo/ncl-talk" target="_blank">http://mailman.ucar.edu/mailman/listinfo/ncl-talk</a><br>
>>><br>
>><br>
><br>
><br>
> _______________________________________________<br>
> ncl-talk mailing list<br>
> List instructions, subscriber options, unsubscribe:<br>
> <a href="http://mailman.ucar.edu/mailman/listinfo/ncl-talk" target="_blank">http://mailman.ucar.edu/mailman/listinfo/ncl-talk</a><br>
><br>
</div></div></blockquote></div></div></div>