[ncl-talk] Issue producing panel plot with single overlay plots

Arne Melsom arne.melsom at met.no
Wed Feb 6 13:52:25 MST 2019 

Hi Anne,
I'm not sure at all if this applies to your application, but I've found
that to produce cropped images on eps output, I can set
 wks_type = "epsi"
...you may wish to rename the resulting output file
 ( mv <file>.epsi <file>.eps )
At least it should be easy enough to test! Good luck.
Arne

Den ons. 6. feb. 2019 kl. 13:16 skrev Rick Brownrigg <brownrig at ucar.edu>:

> Hi Anne,
>
> Its a quirk (feature?) of NCL graphics that plots get mapped onto a
> unit-square plotting space. So plots that have a wide or tall aspect end up
> with margins along their shorter axis.
>
> What most people do is use Imagemagick's "convert" tool with the "-trim"
> option to remove white space along the edges. See this FAQ and search for
> "-trim" for examples.
>
>     http://ncl.ucar.edu/FAQ/
>
> What I don't know offhand is whether "convert" can deal with eps. From the
> examples of the FAQ, it appears that it does work with postscript, so
> hopefully...
>
> Rick
>
> On Wed, Feb 6, 2019 at 2:48 AM Anne <anne.seidenglanz at unive.it> wrote:
>
>> Hi Adam,
>>
>> thanks for taking the time to look into this issue; yes this has indeed
>> worked for me, so I end up with only 1 page instead of 2, great.
>> Remains only unknown to me how to adjust the size of the frame around my
>> plots so as to have a tight frame around the plots. I have used
>>
>>   wks_type = "eps"
>>
>>           wks_type at wkPaperWidthF  = 14.5  ; in inches
>>           wks_type at wkPaperHeightF = 12.0  ; in inches
>>           wks_type at wkWidth = 1500
>>           wks_type at wkHeight = 1500
>>
>> where PaperWidth and PaperHeight change resulted in changing the size of
>> the plots themself (apparently also their resolution), but not the size of
>> the frame around. Is there another resource that I need to specify?
>>
>> Thanks
>> Anne
>>
>> On Tue, Feb 5, 2019 at 9:27 PM Adam Phillips <asphilli at ucar.edu> wrote:
>>
>>> Hi Anne,
>>> Whenever plots are incorrectly being drawn on separate pages or on one
>>> page, it is quite likely that it has to do with incorrect calls to draw or
>>> frame (either implicitly through a plotting function or explicitly).
>>> Calling draw tells NCL to draw the graphic to the workstation, while
>>> calling frame tells NCL to finish the current frame/page and to move to the
>>> next one. gsn_panel and gsn_csm_contour* functions implicitly call both of
>>> those procedures unless the resource list passed in specify otherwise.
>>>
>>> To simplify your script highlighting draw/frame resources and procedures:
>>> res = True
>>> res at gsnDraw = False
>>> res at gsnFrame = False
>>> ...
>>> res2                  = True
>>> res2 at gsnDraw             = False
>>> res2 at gsnFrame            = False
>>> ...
>>> plot(0) = gsn_csm_contour_map(wks, diff_means_lead0,res)
>>> plot0   = gsn_csm_contour(wks, prob_lead0, res2)
>>> overlay(plot(0),plot0)
>>> draw(plot(0))
>>> ; repeat the above 4 lines for plot(1) and plot(2)
>>> frame(wks)
>>>
>>> resP                     = True
>>> resP at gsnPanelXWhiteSpacePercent = 5
>>> gsn_panel(wks, plot, (/1, 3/), resP)
>>>
>>> The above calls draw 3 times, and frame once, before gsn_panel is called
>>> (where gsnFrame/gsnDraw = True). This results in 2 pages in your output
>>> workstation.
>>>
>>> Try this:
>>> res = True
>>> res at gsnDraw = False
>>> res at gsnFrame = False
>>> ...
>>> res2                  = True
>>> res2 at gsnDraw             = False
>>> res2 at gsnFrame            = False
>>> ...
>>> plot(0) = gsn_csm_contour_map(wks, diff_means_lead0,res)
>>> plot0   = gsn_csm_contour(wks, prob_lead0, res2)
>>> overlay(plot(0),plot0)
>>> ; repeat the above 3 lines for plot(1) and plot(2)
>>>
>>> resP                     = True
>>> resP at gsnPanelXWhiteSpacePercent = 5
>>> gsn_panel(wks, plot, (/1, 3/), resP)
>>>
>>> Hope that all makes sense. If you have further questions let ncl-talk
>>> know.
>>> Adam
>>>
>>> On Tue, Feb 5, 2019 at 12:05 PM Anne <anne.seidenglanz at unive.it> wrote:
>>>
>>>> Hi,
>>>>
>>>> I am producing 3 polar projection plots in a panel, whereby each single
>>>> has overlaid contours (using the 'overlay' procedure); see below, I also
>>>> attached the complete script.
>>>> In this way I end up with 2 png files, the 1st showing 1 plot that has
>>>> all 3 plots overlaid on top of each other, and the 2nd one the actual,
>>>> correct, panelled plot. For simple viewing purposes, this is fine for me,
>>>> but when it comes to inserting the plots into a word document, also in
>>>> another format, like .eps format, then only 1 file is being produced and
>>>> when I insert it, only the wrong overlaid 3-in-1 plot appears.
>>>>
>>>> So my question is:  What is the best way in dealing with overlaid(!)
>>>> plots in a panelled configuration, and how to produce them properly (i.e.
>>>> format) if one wants to insert them into word? I also tried to reduce the
>>>> size of the workstation but wasn't successful, they would still have a
>>>> large frame.
>>>>
>>>> My goal is to have a nice, high quality eps file of my (paneled) plot.
>>>>
>>>> Thanks for any help,
>>>> Anne
>>>>
>>>>          res at gsnCenterString = "NDJ "+field+" response"
>>>>         plot(0) = gsn_csm_contour_map(wks, diff_means_lead0, res)
>>>>         plot0   = gsn_csm_contour(wks, prob_lead0, res2)        ; 0.1
>>>> shade all areas less than the
>>>>         plot0   = gsn_contour_shade(plot0, 0.05, 999.,opt)      ; 0.05
>>>> contour level (0.15 for 0.1 contour level)
>>>>
>>>>         overlay(plot(0),plot0)
>>>>         draw(plot(0))
>>>>
>>>>          res at gsnCenterString = "DJF "+field+" response"
>>>>         plot(1) = gsn_csm_contour_map(wks, diff_means_lead1, res)
>>>>         plot1   = gsn_csm_contour(wks, prob_lead1, res2)        ; 0.1
>>>> shade all areas less than the
>>>>         plot1   = gsn_contour_shade(plot1, 0.05, 999.,opt)      ; 0.05
>>>> contour level (0.15 for 0.1 contour level)
>>>>
>>>>         overlay(plot(1),plot1)
>>>>         draw(plot(1))
>>>>
>>>>          res at gsnCenterString = "JFM "+field+" response"
>>>>         plot(2) = gsn_csm_contour_map(wks, diff_means_lead2, res)
>>>>         plot2   = gsn_csm_contour(wks, prob_JFM, res2)  ; 0.1 shade all
>>>> areas less than the
>>>>         plot2   = gsn_contour_shade(plot2, 0.05, 999.,opt)
>>>>    ;0.05 contour level (0.15 for 0.1 contour level)
>>>>
>>>>         overlay(plot(2),plot2)
>>>>         draw(plot(2))
>>>>
>>>>         frame(wks)
>>>>
>>>>
>>>>             resP                     = True
>>>>             resP at gsnPanelXWhiteSpacePercent = 5
>>>>             gsn_panel(wks, plot, (/1, 3/), resP)
>>>> _______________________________________________
>>>> ncl-talk mailing list
>>>> ncl-talk at ucar.edu
>>>> List instructions, subscriber options, unsubscribe:
>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>>
>>>
>>>
>>> --
>>> Adam Phillips
>>> Associate Scientist,  Climate and Global Dynamics Laboratory, NCAR
>>> www.cgd.ucar.edu/staff/asphilli/   303-497-1726
>>>
>>> <http://www.cgd.ucar.edu/staff/asphilli>
>>>
>> _______________________________________________
>> ncl-talk mailing list
>> ncl-talk at ucar.edu
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>
> _______________________________________________
> ncl-talk mailing list
> ncl-talk at ucar.edu
> 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/20190206/ae76ae2c/attachment.html>

More information about the ncl-talk mailing list