[ncl-talk] gsn_csm_streamline_scalar* functions

[Mary Haley]

Hi Alan,

I've added all the gsn_csm_streamline_scalar_xxx functions you contributed
to NCL V6.3.0, and it is also working with paneling.

My version of gsn_panel_return is newer than yours, of course, because I'm
working with the development version, so I made the appropriate changes
based on this.

Using your example code, I got the same plot as you.

I've added variations of your streamline_plot.ncl script to our streamlines
example page.

Thanks again for your contribution and helping me close an old ticket!

​--Mary
​

On Thu, Mar 5, 2015 at 6:58 PM, Alan Brammer <abrammer at albany.edu> wrote:

> Mary,
>
> Having just finished my wild goose chase, here are the answers to your
> questions.
>
> ** Long story short, streamlinePlotClass wasn't supported by gsn_panel. My
> work around caused problems. Here's edited versions of gsn_panel so
> streamlinePlotClass is supported.  **
>
>
> The problem I was having with panelling was because I was drawing one plot
> with lbLabelBarOn = True.  This was throwing off the dimensions and I think
> is a known thing. No big deal I should just use gsnPanelLabelBar = True.
> Well here was the problem, when panelling the supported
> gsn_csm_streamline_map with stMonoLineColor = False, therefore with
> coloured streamlines, gsn_panel says it's an unsupported type and will
> ignore the label bar request.   Hence why I was forcing one of the plots to
> have the label bar.  I don't like having to draw it myself.
>
>
> So, because I wanted the ability to panel my new
> gsn_csm_streamline_scalar* functions as well as the supported
> gsn_csm_streamline_map functions. I've have gone through and edited
> gsn_panel_return and also had to add "streamlinePlotClass"
> to get_plot_not_loglin().
>
>
> Attached below are a simple script to first identify that gsn_panel will
> ignore a labelbar request from the gsn_csm_streamline_map() then load my
> edited versions and show that it now works.
>
>
> Some simple data attached as well.
>
>
>
> On Thu, Mar 5, 2015 at 10:10 AM, Mary Haley <haley at ucar.edu> wrote:
>
>> Hi Alan,
>>
>> Thanks for the code.  Maybe I can get it into V6.3.0.
>>
>> When you say it doesn't panel, what exactly happens? Can you provide a
>> sample test?
>>
>> --Mary
>>
>>
>> On Wed, Mar 4, 2015 at 12:33 PM, Alan Brammer <abrammer at albany.edu>
>> wrote:
>>
>>> I wanted this functionality so made an edited duplicate of the
>>> gsn_csm_vector_scalar* functions.
>>>
>>> Have gone through a tested each function and everything seems to
>>> function without bugs.  Didn't have to edit too much so there wasn't too
>>> much to break.
>>>
>>> gsn_csm_streamline_scalar() doesn't panel as expected for me but this
>>> also happens with me for gsn_csm_vector_scalar() so that's a problem for
>>> another day.
>>>
>>>
>>> I can edit documentation if desired, there is nothing unexpected though.
>>>
>>> If passed wind speed this will function identically to the regular
>>> function with stMonoLineColor = False.  I think this functionality was
>>> available through setting stScalarFieldData, but that didn't seem to match
>>> up with the functionality elsewhere and I'm still not a overly familiar
>>> with setvalues or getvalues.
>>>
>>>
>>> Only tested in 6.2.{0,1}.  I seem to have lost my 6.1 version.
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> ncl-talk mailing list
>>> List instructions, subscriber options, unsubscribe:
>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20150313/2d6afd71/attachment.html