[ncl-talk] issue with dash pattern in multi-panel xy2 plot

[Karin Meier-Fleischer]

Hi Michele,

it is hard to say what is going wrong without the data. But I would 
recommend to create a common resource object which make it easier to 
read and handle the resource settings.

For example

   resCom = True
   resCom at gsnDraw           = False
   resCom at gsnFrame          = False

   resCom at vpWidthF          = 0.70
   resCom at vpHeightF         = 0.175

   resCom at xyLineThicknesses = linethick1
   resCom at xyLineColors      = linecolor1
   resCom at xyDashPatterns    = dashpattern1

   ....

   res4a = resCom
   res4b = resCom


Put all common resources into resCom and only set in res4a and res4b the 
variable resources. You don't need to delete the complete res4a or res4b 
object as it is a error prone. It is better to delete only the 
unwanted/unneeded resources.

Hope this helps. If not, is it possible to get the data file?

-Karin


Am 26.04.21 um 18:44 schrieb Michele Petrini - CITG via ncl-talk:
>
> I am plotting multi-panel timeseries using the gsn_csm_xy2 function, 
> and in each panel I am plotting 6 different curves with different 
> colours/thickness/dash patterns.
>
> Even though the line thickness/dash patterns resources are always the 
> same in each panel, the curves in two panels (second and fourth from 
> the top in the png attached) look different - the lines are slightly 
> thicker, and also the dash pattern is completely different.
>
> I tried to create myself dash patterns, but I got the same difference. 
> I carefully checked all the variables, and they all have same 
> dimensions. One thing that maybe could create some issues is that the 
> time dimension for each variable is quite large (ca. 40,000 values), 
> but still I can’t figure out why this would happen. Do you have any 
> clue why I am getting this and how to fix it?
>
> Please find attached the figure (png), and my ncl code (also, I can 
> provide input netcdf in case you want to replicate the plot).
>
> thanks,
> Michele
>
>
>
>
> ******************************************************************
> Michele Petrini, PhD
>
> Department of Geoscience and Remote Sensing, TUDelft
> Delft, The Netherlands
>
> M.Petrini at tudelft.nl <mailto:M.Petrini at tudelft.nl>
> +39 3398367372
> michele.petrins (Skype alias)
>
>
>
>
> _______________________________________________
> ncl-talk mailing list
> ncl-talk at mailman.ucar.edu
> List instructions, subscriber options, unsubscribe:
> https://mailman.ucar.edu/mailman/listinfo/ncl-talk

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.ucar.edu/pipermail/ncl-talk/attachments/20210427/af68119f/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: timeseries.sensitivity.long.png
Type: image/png
Size: 294191 bytes
Desc: not available
URL: <https://mailman.ucar.edu/pipermail/ncl-talk/attachments/20210427/af68119f/attachment.png>