Re: 20020305: GPFRONT with satellite image

Patrick,

I sent you the ttarfile for gpfront yesterday.

Steve Chiswell

On Wed, 6 Mar 2002, Patrick O'Reilly wrote:

> Steve,
>
> I got the image to plot by first running gpmap and plotting a satellite
> image, then running gpfront and specifying the same satellite image.  If you
> remember back far enough ;) I was getting errors when trying this approach
> before because when I ran gpfront after gpmap, I didn't specify a satellite
> image even though the projection was SAT.  I guess that's where that "Image
> ... does not exist" error came from.
>
>  I would like to get the 5.6.f version of gempak, where you can specify
> LUTFIL in gpfront, but didn't see the binaries for Solaris Sparc online.
> The gpfront you sent me didn't build, as I have the binary version
> installed.  Is the 5.6.f version available yet?  Thanks for your help, I owe
> you one or two.
>
> Patrick
>
> ----- Original Message -----
> From: "Steve Chiswell" <address@hidden>
> To: "Patrick O'Reilly" <address@hidden>
> Cc: <address@hidden>; <address@hidden>
> Sent: Tuesday, March 05, 2002 10:13 PM
> Subject: 20020305: GPFRONT with satellite image
>
>
> >
> >
> > Patrick,
> >
> > It sounds like you don't have enough colors to display the satellite
> > image on your Xvfb screen :1.0. If you have other scripts running
> > and generating their displays on :1.0, especially if they are also
> displaying
> > satellite images, then the possibility exists to have more than one gf
> > process trying to access the color map at the same time.
> > Recall that without ntl running on the display, each process will have its
> own
> > colors, and you could be using 128 satellite colors and 33 graphic colors
> (161
> > colors total).
> >
> > You can try the following:
> >
> > 1) ensure that you don't have other scripts using the :1.0 display at the
> >    time you are running your script. If you are running a lot of scripts
> >    from cron, consider using a master script that queues up the
> >    products that you are producing (also works well to balance the load
> >    on your system).
> >
> > 2) try the gif driver instead of gf. This will avoid the use of the X
> server,
> >    and therefore the color conflicts. If this works, then it likely
> >    means that your X display colors are fewer than necessary.
> >
> > 3) Try launching a second virtual display and using that one.
> >
> > 4) look for any orphaned gf or xw processes using :1.0 that might
> >    have occured if your script exited before running gpend.
> >
> > Steve Chiswell
> > Unidata User SUpport
> >
> >
> >
> >
> >
> > >
> > > Hi Clint,
> > >
> > > Took your script and changed filenames and it worked fine. Funny thing
> > > is.....
> > >
> > > I ran your script from the command line and it worked.  Got the same
> image
> > > you sent me.  I immediately ran the attached script from the command
> line
> > > and got the attached image.  In the attached script, I changed the
> device to
> > > what I would use in a script running through cron.  Also, added a couple
> > > variables for the image name, and set the display to work with the gf
> > > device.  So something's up with that.  I get satellite to plot fine in
> other
> > > scripts using gpmap, but can't get gpfront to plot satellite data with a
> > > script specifying the gf device.
> > >
> > > Another thing, after I ran the attached script and got the attached
> image,
> > > if I run your script again, it no longer works, I get the attached (bad)
> > > image in my x-window.  If I then copy in a default color table, then run
> > > again, your script again works.  In my modified script, I do the same
> thing
> > > I did on the command line, copying in a "black" color table, but no
> dice.
> > > If you have any further ideas, I would love to hear.  Are you confused
> yet?
> > > I am.  Thanks for your help!
> > >
> > > Patrick
> > >
> > >
> > > ----- Original Message -----
> > > From: "Clint Rowe" <address@hidden>
> > > To: <address@hidden>
> > > Cc: <address@hidden>
> > > Sent: Tuesday, March 05, 2002 3:54 PM
> > > Subject: Re: GPFRONT with Satellite
> > >
> > >
> > > > Patrick,
> > > >
> > > > There must be something strange in your script (e.g., a badly defined
> > > > variable).  Try running the extremely simple script I've attached --
> all
> > > > you might need to do is change the filenames.  I get a nice IR image
> with
> > > > fronts overlaid.
> > > >
> > > > Clint
> > > >
> > > >
> > > > >Date: Tue, 05 Mar 2002 12:48:11 -0600
> > > > >From: "Patrick O'Reilly" <address@hidden>
> > > > >Subject: GPFRONT with Satellite
> > > > >To: address@hidden
> > > > >X-MIMEOLE: Produced By Microsoft MimeOLE V5.50.4522.1200
> > > > >X-Priority: 3
> > > > >X-MSMail-priority: Normal
> > > > >
> > > > >Hello,
> > > > >
> > > > >I am having lots of trouble getting a plot of fronts with IR
> satellite
> > > using
> > > > GPFRONT in a csh script.  I can get the plot when sitting at a
> terminal
> > > running
> > > > gempak interactively, but when I use a script, strange things happen.
> I
> > > am not
> > > > running the script from cron, but executing it as user gempak from the
> > > command
> > > > line.  In my scripts, I have tried using just GPFRONT with a satellite
> > > image
> > > > specified. In this case I get an image of fronts, the IR satellite
> color
> > > bar
> > > > legend, but no satellite image.   I have also tried first plotting a
> > > satellite
> > > > image using GPMAP , then running GPFRONT to overlay the fronts.  In
> this
> > > case, I
> > > > get a satellite image, but no fronts drawn and an error: "[IM -1]
> Image
> > > file ...
> > > > does not exist."  Does anyone have a script running that produces this
> > > image (IR
> > > > and fronts)?  If so, I would love to see a copy.  I have about 20
> scripts
> > > with
> > > > exactly the same "setup" running successfully, creating surface,
> model,
> > > > satellite, and radar images and overlays.  This one won't work,
> though.
> > > Any
> > > > input would be appreciated. Thanks.
> > > > >
> > > > >Patrick
> > > > >
> > > > >~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> > > > >Patrick O'Reilly                               Support Scientist
> > > > >The STORM Project            address@hidden
> > > > >208 Latham Hall                             ph: 319-273-3789
> > > > >University of Northern Iowa
> > > > >Cedar Falls, IA 50614
> > > > >~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> > > > >
> > > >
> > > >
> > > > ====================================================================
> > > > Clinton M. Rowe
> > > > Associate Professor
> > > > Meteorology/Climatology Program                  phone:(402)472-1946
> > > > Department of Geosciences                          fax:(402)472-4917
> > > > University of Nebraska-Lincoln                        address@hidden
> > > >
> > > >
> > >
> >
>