[ncl-talk] getting empty plots

[Brian Thomas]

Turns out that upgrading to 6.6.2 fixed the issue.  I used the conda install method, which worked fine.  The tests after install worked and my scripts are now producing good plots.  Thanks for the help!

On Feb 11, 2020, at 2:34 PM, Rick Brownrigg <brownrig at ucar.edu<mailto:brownrig at ucar.edu>> wrote:

Brian,

We had seen this issue occasionally a few years back with systems that had builtin Intel graphics, and most often on Ubuntu systems (I was never convinced it had anything to do with Ubuntu, and everything to do with the Intel graphics drivers). The behavior is part of a plot gets drawn, but it's as though the last "buffer worth" does not get flushed. And this only happens when drawing to an X11 window -- all other file-based workstation outputs are OK. There is note this in the FAQ:

When I send my NCL graphics to an X11 window, only a partial image is drawn.

We have seen issues with X11 on machines that have embedded Intel graphics processors. The behavior is that the plot is only partially drawn, and if you run the script several times back-to-back you may see the plot is drawn to varying/random levels of (in)completeness.

If you determine you have an Intel GPU on your machine, make sure you have their latest graphics drivers installed.


This issue hasn't been reported very much in more recent years. Since your machine has an NVIDIA GPU, I don't know what to recommend. The default out-of-the-box, open source graphics driver for NVIDIA GPU's is nouveau, and I have to think that's what most of our users go with. NVIDIA offers its own proprietary driver, but it takes considerably more awareness to i) install it, and ii) disable the nouveau driver.  If you feel up to the task, that might be a place to start.  I'd also note that the latest NCL rev is v6.2.2 -- perhaps its worth upgrading to that.

I hope that helps and wish I had a better answer...

Rick


On Tue, Feb 11, 2020 at 12:56 PM Brian Thomas <brian.thomas at washburn.edu<mailto:brian.thomas at washburn.edu>> wrote:
I’m using Mint 19.3, which is an Ubuntu derivative.  I was on openSUSE before.  I have Intel CPUs but an Nvidia graphics card.  The NCL version is 6.4.0

On Feb 11, 2020, at 1:12 PM, Rick Brownrigg <brownrig at ucar.edu<mailto:brownrig at ucar.edu>> wrote:

Are you by chance now using Ubuntu and a machine with builtin Intel graphics processor (GPU). Otherwise, what flavors of linux did you change from/to, and which version of ncl (ncl -v)?



On Tue, Feb 11, 2020 at 12:06 PM Brian Thomas via ncl-talk <ncl-talk at ucar.edu<mailto:ncl-talk at ucar.edu>> wrote:
I’ve been using NCL for a while and been able to do a lot with it, but I recently changed linux OS flavors and just discovered that my scripts are now producing empty plots.  This is the case for different data and different scripts.  I’ve checked data and coordinates to be sure they are ok.

The plot frame, labels and color bars are drawn, but the inside of the plot is empty.

I’ve tried removing and re-installing NCL, which didn’t help.

Has anyone else seen anything like this?

Thanks!

@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@
Brian C. Thomas, PhD
Professor, Physics and Astronomy, Washburn University

785-670-2144<tel:(785)%20670-2144>
brian.thomas at washburn.edu<mailto:brian.thomas at washburn.edu>

Web: http://www.washburn.edu/faculty/bthomas/
ORCID profile: http://orcid.org/0000-0001-9091-0830
ResearchGate profile: https://www.researchgate.net/profile/Brian_Thomas

Department of Physics and Astronomy
Washburn University
1700 SW College Ave.<https://maps.google.com/?q=1700+SW+College+Ave.%0D+Topeka,+KS+66621&entry=gmail&source=g>
Topeka, KS 66621<https://maps.google.com/?q=1700+SW+College+Ave.%0D+Topeka,+KS+66621&entry=gmail&source=g>

@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@

_______________________________________________
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

@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@
Brian C. Thomas, PhD
Professor, Physics and Astronomy, Washburn University

785-670-2144<tel:(785)%20670-2144>
brian.thomas at washburn.edu<mailto:brian.thomas at washburn.edu>

Web: http://www.washburn.edu/faculty/bthomas/
ORCID profile: http://orcid.org/0000-0001-9091-0830
ResearchGate profile: https://www.researchgate.net/profile/Brian_Thomas

Department of Physics and Astronomy
Washburn University
1700 SW College Ave.<https://maps.google.com/?q=1700+SW+College+Ave.%0D+Topeka,+KS+66621&entry=gmail&source=g>
Topeka, KS 66621<https://maps.google.com/?q=1700+SW+College+Ave.%0D+Topeka,+KS+66621&entry=gmail&source=g>

@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@


@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@
Brian C. Thomas, PhD
Professor, Physics and Astronomy, Washburn University

785-670-2144<tel:(785)%20670-2144>
brian.thomas at washburn.edu<mailto:brian.thomas at washburn.edu>

Web: http://www.washburn.edu/faculty/bthomas/
ORCID profile: http://orcid.org/0000-0001-9091-0830
ResearchGate profile: https://www.researchgate.net/profile/Brian_Thomas

Department of Physics and Astronomy
Washburn University
1700 SW College Ave.<https://maps.google.com/?q=1700+SW+College+Ave.%0D+Topeka,+KS+66621&entry=gmail&source=g>
Topeka, KS 66621<https://maps.google.com/?q=1700+SW+College+Ave.%0D+Topeka,+KS+66621&entry=gmail&source=g>

@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@#@

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20200212/06597cc9/attachment.html>