[ncl-talk] weird color shown in plots

Ming Chen chenming at ucar.edu
Wed Apr 10 10:04:13 MDT 2019


Rick,

I do have the .hluresfile in my home directory. After commenting out the 
wkColorMap, I got the correct color.

Thank you !

Ming

On 4/10/19 9:54 AM, Rick Brownrigg wrote:
> Hi Ming,
>
> I just tested this from home, and I can't reproduce the issue. I 
> copied the NCL script directly from the directory you cited in the 
> original email. I had it write to both X11 and PNG workstations, and 
> I'm getting the correct colors. I simply do a "module load ncl" upon 
> logging in to cheyenne.
>
> Something just occurred to me -- do you by chance have a file name 
> ".hluresfile" (note the leading ".") in your home directory, and does 
> it have an entry for "*wkColorMap"?  (Note if its prefixed with a "!", 
> its commented out)
>
> Rick
>
>
> On Wed, Apr 10, 2019 at 9:35 AM Ming Chen <chenming at ucar.edu 
> <mailto:chenming at ucar.edu>> wrote:
>
>     Rick,
>
>     I turned off all the default settings. I only loaded intel
>     compiler and ncl/6.5.0, But I still got the wrong color.
>
>     Is this possibly an issue of environmental settings?
>
>     Ming
>
>     On 4/9/19 10:06 AM, Rick Brownrigg wrote:
>>     Well, now I can't reproduce the issue!   I saw what you were
>>     seeing last night while running on cheyenne logged in from home.
>>     Now this morning, logged into cheyenne from the office, I'm
>>     seeing the expected colors. This does not make sense! I will test
>>     again from home this evening, but I have no explanation as to why
>>     that should make any difference :-/
>>
>>     More as I know it.
>>
>>
>>     On Tue, Apr 9, 2019 at 9:48 AM Ming Chen <chenming at ucar.edu
>>     <mailto:chenming at ucar.edu>> wrote:
>>
>>         Rick,
>>
>>         Thank you for getting back to me quickly. I just type 'module
>>         load ncl' to enable ncl in cheyenne.   The command "module
>>         list" shows
>>
>>         Currently Loaded Modules:
>>           1) ncarenv/1.2    3) ncarcompilers/0.4.1   5)
>>         netcdf/4.6.1   7) nco/4.7.4   9) ncview/2.1.7
>>           2) intel/17.0.1   4) mpt/2.19              6)
>>         xxdiff/4.0.1   8) ncl/6.5.0  10) grib-bins/1.1
>>
>>         The command  "echo $NCARG_ROOT"  yields
>>         "/glade/u/apps/ch/opt/ncl/6.5.0/intel/17.0.1"
>>
>>         I notice this issue because the plots in my desktop and the
>>         plots in cheyenne are different when I run the same NCL
>>         script and process the same data.
>>
>>         Thanks.
>>
>>         Ming
>>
>>
>>         On 4/8/19 7:10 PM, Rick Brownrigg wrote:
>>>         Hi Ming,
>>>
>>>         This is very strange. I see the differences in color that
>>>         you refer to, but I downloaded the ncl script from your
>>>         directory and I get the same colors as shown on the web
>>>         site. I see that there's no attempt in the script to modify
>>>         the colormap, so it should be the default colormap
>>>         (http://ncl.ucar.edu/Document/Graphics/ColorTables/ncl_default.shtml),
>>>         However your plots have a few colors that are not part of
>>>         that default map (??)
>>>
>>>         How are you enabling NCL in cheyenne, and what does "echo
>>>         $NCARG_ROOT" report?
>>>
>>>         Rick
>>>
>>>
>>>
>>>
>>>         On Mon, Apr 8, 2019 at 5:24 PM Ming Chen <chenming at ucar.edu
>>>         <mailto:chenming at ucar.edu>> wrote:
>>>
>>>             Hello,
>>>
>>>             I found that the colors in my plot are not the colors I
>>>             specify. Please
>>>             see the files saved in /glade/scratch/chenming/station/
>>>
>>>             I downloaded "station_2.ncl"  from NCL website, and run
>>>             it without any
>>>             modification. The png files are quite different to the
>>>             sample figures
>>>             shown in the NCL website.  In my case, it seems that
>>>             somehow the colors
>>>             are changed.
>>>
>>>             I run this  script on cheyenne. Does anyone know what
>>>             could be the reason?
>>>
>>>             Thanks.
>>>
>>>             Ming
>>>
>>>
>>>
>>>             _______________________________________________
>>>             ncl-talk mailing list
>>>             ncl-talk at ucar.edu <mailto: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/20190410/0add8191/attachment.html>


More information about the ncl-talk mailing list