Re: ncl-talk Digest, Vol 123, Issue 33

From: Lijun Diao <ljdiao_at_nyahnyahspammersnyahnyah> Date: Fri Feb 28 2014 - 12:18:28 MST · >

-- 
Dipl. Geophys. Karin Meier-Fleischer
Application Support, Visualization
Deutsches Klimarechenzentrum GmbH      E-Mail:   meier-fleischer@dkrz.de
Bundesstrasse 45a                      Internet: http://www.dkrz.de/
20146 Hamburg                          Phone:    +49 (0)40 460094 126
Germany                                Fax:      +49 (0)40 460094 270
Gesch?ftsf?hrer: Prof. Dr. Thomas Ludwig
Sitz der Gesellschaft: Hamburg
Amtsgericht Hamburg HRB 39784
-------------- next part --------------
An HTML attachment was scrubbed...
URL: 
http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20140228/f6e135ba/attachment.html
------------------------------
Message: 10
Date: Fri, 28 Feb 2014 16:41:24 +0000
From: Ciara O'Hara <cohara1985@hotmail.com>
Subject: looping to create a graph from many different
files
To: "ncl-talk@ucar.edu" <ncl-talk@ucar.edu>
Message-ID: <DUB111-W103A3C97E8B61C403DED054DC820@phx.gbl>
Content-Type: text/plain; charset="iso-8859-1"
        I have 50 files that I want to run a script on one after the
other, and to save the generated graph with a unique name each time. My
script to create the graph is fine, but looping through the 50 files is
not. I have left out many of the resources I'm using.
Thanks in advance,
Ciara
My script is:
    load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl"
    load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_csm.ncl"
    load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/contributed.ncl"
    load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/shea_util.ncl"
    do n=1961,2010
        begin
        fnam="/home/cohara/TempData/yearly_data/average/average_" + 
sprinti("0.4n",n) + ".nc"
        x=addfile(fnam,"r")
        data=x->var61(0,0,:,:)
        xwks=gsn_open_wks("ps","Average_" + sprinti("0.4n",n)
        resources=True
        resources@tiMainString="Average Annual Temperature" + 
sprinti("0.4n",n)
            plot=gsn_csm_contour_map(xwks,data,resources)
        end
    end do
-------------- next part --------------
An HTML attachment was scrubbed...
URL: 
http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20140228/4b139854/attachment.html
------------------------------
Message: 11
Date: Fri, 28 Feb 2014 09:57:24 -0700
From: Craig Tierney - NOAA Affiliate <craig.tierney@noaa.gov>
Subject: Re: Why doesn't my text show up in my plot?
To: Mary Haley <haley@ucar.edu>
Cc: ncl-talk@ucar.edu
Message-ID:
<CALQst9gxZvaN59EBTHrTT9St4bX2qM1_WK_DD8pM2p0FmTpwAA@mail.gmail.com>
Content-Type: text/plain; charset="iso-8859-1"
Mary,
In developing an example I found the problem, but it lead to another issue.
I had not set res@gsnFrame and res@gsnDraw to false like I had done in
other similar commands, so I was unable to plot any additional objects
after the call to gsn_csm_contour_map.  So that is pretty obvious why it
isn't working.
However, this lead to another issue I had never asked about. I would rather
be generating png files directly, and not creating ps files and converting
them after the fact.  This code was able to write directly to png files
with:
wks=gsn_open_wks("png",ofn);
However, when I changed gsnFrame and gsnDraw to false, I could no longer
create png files.  I can create ps files by just changing the command above
to use "ps".
Should I be able to create png files directly with gsnFrame and gsnDraw set
to false?
If you need to see my example, please send me instructions on where to
upload it.
Thanks,
Craig
On Wed, Feb 26, 2014 at 3:06 PM, Mary Haley <haley@ucar.edu> wrote:
>
> On Feb 26, 2014, at 1:54 PM, Craig Tierney - NOAA Affiliate <
> craig.tierney@noaa.gov> wrote:
>
> > Mary,
> >
> > I will try and come up with an example.  The gsn_add_text isn't working
> either.  Now, I could be using it wrong (most likely).
> >
> > My question is, should gsn_add_text be able to plot text outside of the
> boundary of the plot map?
>
> No, you can only use gsn_add_text for text inside the plot's boundaries,
> although you can do a kludgy thing of moving the text by adding extra
> spaces or carriage returns to the text string.
>
> --Mary
>
> >
> > Thanks,
> > Craig
> >
> >
> > On Wed, Feb 26, 2014 at 1:26 PM, Craig Tierney - NOAA Affiliate <
> craig.tierney@noaa.gov> wrote:
> > On Wed, Feb 26, 2014 at 12:45 PM, Mary Haley <haley@ucar.edu> wrote:
> > Craig,
> >
> > Can you trim this down to a short but complete script that illustrates
> the problem? I can't tell from looking at the code snippet what might be
> going on.
> >
> >
> > I will see what I can do.  The plotting script is really 1000+ lines of
> NCL that allow me plot any dataset, from any resolution, of any
> meteorological field, and just do the right thing.  This script is 
> plotting
> a data on a icosahederal grid.
> >
> > There could be a maximization issue, or maybe one plot element is
> covering up another, etc.
> >
> > Note: generally I recommend against using gsn_text_ndc (or any of the
> gsn_xxx_ndc procedures) if you can use gsn_add_xxx instead.
> >
> >
> > I was not a fan of working in NDC space, but it did help me simplify my
> code because when the scripts change domain (eg. global to conus), then I
> would have to rescale the coordinates.  In NDC space and an assumption of
> where the plot ends up, I can safely tuck away the text in the corner of
> the plot.
> >
> > The difference, of course, is that with gsn_xxx_ndc you need to use NDC
> coordinates (coordinates from 0 to 1), whereas with the gsn_add_xxx
> functions  you need to use the plot's coordinate space. Also, with
> gsn_xxx_ndc, the primitive gets drawn right in that moment and it is *not*
> attached to anything. The gsn_add_xxx function will attach the primitive 
> to
> the plot, but not draw it. So, you have to draw the plot in order to see
> the primitive. But, the benefit is if you resize the plot, then the
> primitive will be resized automatically.
> >
> >
> > I will try the gsn_add_text and see if that resolves the order issue.
> >
> > Craig
> > --Mary
> >
> > On Feb 26, 2014, at 11:59 AM, Craig Tierney - NOAA Affiliate <
> craig.tierney@noaa.gov> wrote:
> >
> > > I am using ncl 6.1.2.  When I try to call gsn_text_ndc before the call
> to gsn_csm_contour_map, the text displays.
> > >
> > > ----------------------------
> > > ; Create a control string for the plot
> > > cstr="Min/Max control: "+min(d1)+"/"+max(d1)+" "+"Min/Max Diff:
> "+min(data)+"/"+max(data)
> > >  print(cstr)
> > >
> > > ; Attach all necessary labels
> > > txresb   = True
> > > txresb@txFontHeightF = 0.010
> > > gsn_text_ndc(wks,cstr, 0.5, 0.1, txresb)
> > >
> > >  ; plot the map
> > > map = gsn_csm_contour_map(wks,data,res)
> > >
> > > draw(map)
> > > -----------------------------
> > >
> > > When I reverse it, the text does not display.
> > >
> > > ----------------------------
> > > ; Create a control string for the plot
> > > cstr="Min/Max control: "+min(d1)+"/"+max(d1)+" "+"Min/Max Diff:
> "+min(data)+"/"+max(data)
> > >  print(cstr)
> > >
> > >  ; plot the map
> > > map = gsn_csm_contour_map(wks,data,res)
> > >
> > > ; Attach all necessary labels
> > > txresb   = True
> > > txresb@txFontHeightF = 0.010
> > > gsn_text_ndc(wks,cstr, 0.5, 0.1, txresb)
> > >
> > > draw(map)
> > > -----------------------------
> > >
> > > Why is this and is there a way to enable additional debugging so that
> I could see what NCL is doing when I run into future problems like this?
> > >
> > > Thanks,
> > > Craig
> > > _______________________________________________
> > > 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/20140228/f648816e/attachment.html
------------------------------
Message: 12
Date: Fri, 28 Feb 2014 10:37:10 -0700
From: Mary Haley <haley@ucar.edu>
Subject: Re: Adding polygons to a Skew-T plot
To: Katharina Roloff <roloff@muk.uni-hannover.de>
Cc: ncl-talk@ucar.edu
Message-ID: <BAF3A791-1B49-44DC-9ADF-FCD61F5A8437@ucar.edu>
Content-Type: text/plain; charset="iso-8859-1"
Hi,
gsn_add_polygon is the correct route to go, and not gsn_add_annotation.
It looks like you are very close, but the X/Y values you're using for the 
polygon are not in the same coordinate space as the Skew-T plot.
You need to use the unadvertised "skewtx" and "skewty" functions to convert 
from your coordinates to the Skew-T coordinates:
      ybox = 
skewty((/p_data(0,frz_base),p_data(0,frz_top),p_data(0,frz_top),p_data(0,frz_base),p_data(0,frz_base)/))
      xbox = skewtx((/30.,30.,37.,37.,30./),ybox)
      box_frz = gsn_add_polygon (wks,plot,xbox,ybox,boxres)
Dennis is more familiar with this code, so I'm hoping he'll confirm whether 
this is correct.
You can see the skewtx and skewty functions for yourself in:
$NCARG_ROOT/lib/ncarg/nclscripts/csm/skewt_func.ncl
Search for "function skewty" and "function skewtx":
undef("skewty")
function skewty(pres[*]:numeric)            ; y-coord given pressure (mb)
  begin
   return (132.182-44.061*log10(pres))
  end
undef("skewtx")
function skewtx(temp[*]:numeric,y[*]:numeric) ; x-coord given temperature 
(c)
  begin                                     ; y=skewty(p)
   return (0.54*temp+0.90692*y)
  end
--Mary
On Feb 27, 2014, at 10:42 AM, Katharina Roloff <roloff@muk.uni-hannover.de> 
wrote:
> Hi all,
>
> after trying to panel some Skew-T plots last week I've now tried to add a 
> polygon to a single Skew-T plot similar to an example of the NCL talk 
> archive
>
> http://www.ncl.ucar.edu/Support/talk_archives/2013/1742.html
>
> but unfortunately it doesn't work at all. Using the gsn_add_annotation 
> function I get the error messages
>
> fatal:NhlAddAnnotation: invalid view id
> fatal:["NclVar.c":1382]:Assignment type mismatch, right hand side can't be 
> coerced to type of left hand side
> fatal:["Execute.c":8128]:Execute: Error occurred at or near line 186 in 
> file Skew_T_Scenarios.ncl
>
> When I just try to add the polygon I can't see it after plotting. Maybe 
> you have any idea what's going wrong? I don't know if the polygon is 
> hidden by the Skew-T background even though I tried different options to 
> make sure that the polygon is plotted last...
>
> In the attachment you see the shortened code and the ps output which just 
> shows the Skew-T plot without any polygon.
>
> If you need the input files, I've sent them via ftp in your /incoming 
> directory:
>
> cosmo_tools.ncl
> grid_calculation.ncl
> grb1_192_pip_LME_sce_k_2013110600_04_ueg
> grb12-lm2mo_scal_2013110600_04.ueg
> mySkewT.ncl
>
> I am working with NCL 6.1.2. When I type "uname -a" I get
>
>> Linux 3.1.10-1.16-desktop #1 SMP PREEMPT Wed Jun 27 05:21:40 UTC 2012 
>> (d016078) x86_64 x86_64 x86_64 GNU/Linux
>
> I would appreciate any assistance again. Thank you, Katharina.
>
>
> --------------------------------------------
>
> M.Sc. Katharina Roloff
> Wissenschaftliche Mitarbeiterin
> Institut fuer Meteorologie und Klimatologie
> Leibniz Universitaet Hannover
> Herrenhaeuser Str. 2
> 30419 Hannover
>
> Email: roloff@muk.uni-hannover.de
>
> --------------------------------------------
> <Skew_T_Scenarios.ncl><Skew_T_frz_20131106_0400.ps>_______________________________________________
> 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/20140228/fa9476c2/attachment.html
------------------------------
Message: 13
Date: Fri, 28 Feb 2014 10:48:00 -0700
From: Mary Haley <haley@ucar.edu>
Subject: Re: Why doesn't my text show up in my plot?
To: Craig Tierney - NOAA Affiliate <craig.tierney@noaa.gov>
Cc: ncl-talk@ucar.edu
Message-ID: <5FE8A4B8-0C80-4FB6-80E7-75C593E8A4EE@ucar.edu>
Content-Type: text/plain; charset="windows-1252"
Craig,
If you set gsnFrame and gsnDraw to False, then the plot object will be 
created, but nothing will get drawn. It shouldn't matter if your output is 
going to PS, PNG, X11, PDF, or NCGM.
Whenever you set these two resources to False, then in order to have the 
plot drawn correct, you need to call:
  draw(plot)
  frame(wks)
where "plot" is whatever you used for the variable on the left side of the 
"=":
   plot = gsn_csm_contour_map(?)
I think what's happening in your case is that if you set these two resources 
to False, and don't call "draw" and "frame", then for the "ps" case, you 
will still get a "xxxx.ps" file, but it will be blank. For the "png" case, 
you will get no file at all.
Look at your PS file to see if it actually has any graphics in it.  I won't 
be surprised if it doesn't.
Are you actually calling "draw" and "frame" somewhere in your code?
See the attached script for an example. Try commenting out the "draw" and 
"frame" calls and see what happens.
--Mary
-------------- next part --------------
A non-text attachment was scrubbed...
Name: map.ncl
Type: application/octet-stream
Size: 344 bytes
Desc: not available
Url : 
http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20140228/e3e1cd16/attachment.obj
-------------- next part --------------
On Feb 28, 2014, at 9:57 AM, Craig Tierney - NOAA Affiliate 
<craig.tierney@noaa.gov> wrote:
> Mary,
>
> In developing an example I found the problem, but it lead to another 
> issue.  I had not set res@gsnFrame and res@gsnDraw to false like I had 
> done in other similar commands, so I was unable to plot any additional 
> objects after the call to gsn_csm_contour_map.  So that is pretty obvious 
> why it isn't working.
>
> However, this lead to another issue I had never asked about. I would 
> rather be generating png files directly, and not creating ps files and 
> converting them after the fact.  This code was able to write directly to 
> png files with:
>
> wks=gsn_open_wks("png",ofn);
>
> However, when I changed gsnFrame and gsnDraw to false, I could no longer 
> create png files.  I can create ps files by just changing the command 
> above to use "ps".
>
> Should I be able to create png files directly with gsnFrame and gsnDraw 
> set to false?
>
> If you need to see my example, please send me instructions on where to 
> upload it.
>
> Thanks,
> Craig
>
>
>
> On Wed, Feb 26, 2014 at 3:06 PM, Mary Haley <haley@ucar.edu> wrote:
>
> On Feb 26, 2014, at 1:54 PM, Craig Tierney - NOAA Affiliate 
> <craig.tierney@noaa.gov> wrote:
>
> > Mary,
> >
> > I will try and come up with an example.  The gsn_add_text isn't working 
> > either.  Now, I could be using it wrong (most likely).
> >
> > My question is, should gsn_add_text be able to plot text outside of the 
> > boundary of the plot map?
>
> No, you can only use gsn_add_text for text inside the plot's boundaries, 
> although you can do a kludgy thing of moving the text by adding extra 
> spaces or carriage returns to the text string.
>
> --Mary
>
> >
> > Thanks,
> > Craig
> >
> >
> > On Wed, Feb 26, 2014 at 1:26 PM, Craig Tierney - NOAA Affiliate 
> > <craig.tierney@noaa.gov> wrote:
> > On Wed, Feb 26, 2014 at 12:45 PM, Mary Haley <haley@ucar.edu> wrote:
> > Craig,
> >
> > Can you trim this down to a short but complete script that illustrates 
> > the problem? I can't tell from looking at the code snippet what might be 
> > going on.
> >
> >
> > I will see what I can do.  The plotting script is really 1000+ lines of 
> > NCL that allow me plot any dataset, from any resolution, of any 
> > meteorological field, and just do the right thing.  This script is 
> > plotting a data on a icosahederal grid.
> >
> > There could be a maximization issue, or maybe one plot element is 
> > covering up another, etc.
> >
> > Note: generally I recommend against using gsn_text_ndc (or any of the 
> > gsn_xxx_ndc procedures) if you can use gsn_add_xxx instead.
> >
> >
> > I was not a fan of working in NDC space, but it did help me simplify my 
> > code because when the scripts change domain (eg. global to conus), then 
> > I would have to rescale the coordinates.  In NDC space and an assumption 
> > of where the plot ends up, I can safely tuck away the text in the corner 
> > of the plot.
> >
> > The difference, of course, is that with gsn_xxx_ndc you need to use NDC 
> > coordinates (coordinates from 0 to 1), whereas with the gsn_add_xxx 
> > functions  you need to use the plot's coordinate space. Also, with 
> > gsn_xxx_ndc, the primitive gets drawn right in that moment and it is 
> > *not* attached to anything. The gsn_add_xxx function will attach the 
> > primitive to the plot, but not draw it. So, you have to draw the plot in 
> > order to see the primitive. But, the benefit is if you resize the plot, 
> > then the primitive will be resized automatically.
> >
> >
> > I will try the gsn_add_text and see if that resolves the order issue.
> >
> > Craig
> > --Mary
> >
> > On Feb 26, 2014, at 11:59 AM, Craig Tierney - NOAA Affiliate 
> > <craig.tierney@noaa.gov> wrote:
> >
> > > I am using ncl 6.1.2.  When I try to call gsn_text_ndc before the call 
> > > to gsn_csm_contour_map, the text displays.
> > >
> > > ----------------------------
> > > ; Create a control string for the plot
> > > cstr="Min/Max control: "+min(d1)+"/"+max(d1)+" "+"Min/Max Diff: 
> > > "+min(data)+"/"+max(data)
> > >  print(cstr)
> > >
> > > ; Attach all necessary labels
> > > txresb   = True
> > > txresb@txFontHeightF = 0.010
> > > gsn_text_ndc(wks,cstr, 0.5, 0.1, txresb)
> > >
> > >  ; plot the map
> > > map = gsn_csm_contour_map(wks,data,res)
> > >
> > > draw(map)
> > > -----------------------------
> > >
> > > When I reverse it, the text does not display.
> > >
> > > ----------------------------
> > > ; Create a control string for the plot
> > > cstr="Min/Max control: "+min(d1)+"/"+max(d1)+" "+"Min/Max Diff: 
> > > "+min(data)+"/"+max(data)
> > >  print(cstr)
> > >
> > >  ; plot the map
> > > map = gsn_csm_contour_map(wks,data,res)
> > >
> > > ; Attach all necessary labels
> > > txresb   = True
> > > txresb@txFontHeightF = 0.010
> > > gsn_text_ndc(wks,cstr, 0.5, 0.1, txresb)
> > >
> > > draw(map)
> > > -----------------------------
> > >
> > > Why is this and is there a way to enable additional debugging so that 
> > > I could see what NCL is doing when I run into future problems like 
> > > this?
> > >
> > > Thanks,
> > > Craig
> > > _______________________________________________
> > > ncl-talk mailing list
> > > List instructions, subscriber options, unsubscribe:
> > > http://mailman.ucar.edu/mailman/listinfo/ncl-talk
> >
> >
> >
>
>
------------------------------
_______________________________________________
ncl-talk mailing list
ncl-talk@ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
End of ncl-talk Digest, Vol 123, Issue 33
***************************************** 
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk