[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 20050117: Ice bulb temperature



Patrick,

Now you aren't getting complaints about not finding TMIK,
so we now that the structure of how you have added the
routine is now correct. You must provide the three grids
to the functioon, as these are the grids that are getting popped
of the stack as grids num1, num2, and num3 in your
dftmik routine.

Presumably, the grids are the correct order. Your
routine is calling PR_TMIK. Does this routine iterate
for every point eg i=1,npt for the kxyd that is passed?

Some of the routines are set up to call a PD_xxxx routine, that iterates
for every point ans calls a pr_xxxx routine, which is the pr routine
that a single point is calculated such as the sfmap program would call.

Since you values show up now as min and max of zero,
where your SCALE=0, you might try setting SCALE= just to see if
there are data values there. Otherwise, they might all be missing
(-9999).

You are probably at the point where a good test would be to put a write
statement in your pr_tmik and watch the values that are being
calculated.

Steve Chiswell
Unidata User Support





On Tue, 2005-03-29 at 12:41, Patrick S. Market wrote:
> > 
> > Patrick,
> > 
> > You didn't show your GFUNC setting in gdcntr, but presumably you have:
> > GFUNC=TMIK(PRES,TMPK,RMIX)
> 
> I have done it with and without.  When I use the whole line, I
> get a bit further, with an error of "NO CONTOUR LEVELS COMPUTED":
> 
>  GEMPAK-GDCNTR>li
>  GDATTIM  = 970410/00f00
>  GLEVEL   = 0
>  GVCORD   = none
>  GFUNC    = tmik(pres,tmpk,mixr)
>  GDFILE   = 25kmFP.gem
>  CINT     =  
>  LINE     = 1/2/3
>  MAP      = 1/1/2
>  TITLE    = 1/-3
>  DEVICE   = xw
>  SATFIL   =  
>  RADFIL   =  
>  IMCBAR   =  
>  PROJ     = str/90.0;-90.70;0.0
>  GAREA    = stl*
>  IJSKIP   =  
>  CLEAR    = yes
>  PANEL    = 0
>  TEXT     = 1/hw/1
>  SCALE    = 0
>  LATLON   = 0
>  HILO     =  
>  HLSYM    =  
>  CLRBAR   = 1/H/LR/0.95;.02/0.2
>  CONTUR   = 4/2
>  SKIP     = 0
>  FINT     = -1;0;1
>  FLINE    = 0;21;21;0
>  CTYPE    = c
>  LUTFIL   =  
>  STNPLT   =  
>  GEMPAK-GDCNTR>dev=xw
>  GEMPAK-GDCNTR>r
> 
> 
> Creating process: xw for queue 405
> 
>  Grid file: 25kmFP.gem                                                        
>                                                               
>  GRID IDENTIFIER: 
>     TIME1             TIME2         LEVL1 LEVL2   VCORD PARM
> 970410/0000F000                         0          NONE TMIKPRESTMPK
> 
>  GAREA:     stl*                                                              
>                                                                SCALE:  0
> 
>  MINIMUM AND MAXIMUM VALUES     0.00     0.00
> NO CONTOUR LEVELS COMPUTED
> Enter <cr> to accept parameters or type EXIT: [GEMPLT -5]  NOMFIL - Invalid 
> map file defined -- default used.
>  Parameters requested: GDATTIM,GLEVEL,GVCORD,GFUNC,GDFILE,CINT,LINE,MAP,
>  TITLE,DEVICE,SATFIL,RADFIL,IMCBAR,PROJ,GAREA,IJSKIP,CLEAR,PANEL,TEXT,SCALE,
>  LATLON,HILO,HLSYM,CLRBAR,CONTUR,SKIP,FINT,FLINE,CTYPE,LUTFIL,STNPLT.
>  GEMPAK-GDCNTR> GEMPAK-GDCNTR>
> 
> 
> 
> 
> > 
> > Double check your dgdriv.f routine to verify that your
> > grid driver routine knows about your df_tmik routine?
> 
> did it.
> 
> > 
> > Now that we know your dftmik.o routine is in gemlib.a,
> > and if you have added the call to that routine in dgdriv.f
> > so that it knows to call your routine when it is given the TMIK
> > function- you can verify that this routine has been updated in the
> > library with:
> > ar tv $GEBLIB/gemlib.a dgdriv.o
> 
> Isentropic Highlander =>ar tv $GEMLIB/gemlib.a dgdriv.o
> rw-rw-r--  1001/    10  27220 Mar 24 09:24 2005 dgdriv.o
> 
> 
> > 
> > After you make any changes to the library routines, you need to relink
> > your GD programs such as gdcntr which you show below with:
> > 
> > cd $GEMPAK/source/programs/gd
> > make clean
> > make all
> > make install
> > make clean
> > 
> > And similarly for NMAP2 and Garp to recognize your routine:
> > 
> > cd $GEMPAK/source/programs/gui
> > make clean
> > make all
> > make install
> > make clean
> > 
> > cd $GARPHOME
> > make clean
> > make all
> > make install
> > make clean
> > 
> 
> did it all, just to be thorough.  I still get "NO CONTOUR LEVELS COMPUTE"
> on TMIK (the output above was done after all of this, actually),
> although TMWK works just fine.  By the way, I do have grids of
> PRES, TMPK, and MIXR at 0%NONE.
> 
> Pat
> 
> > 
> > Steve Chiswell
> > Unidata User Support        
> > 
> > 
> > 
> > On Tue, 2005-03-29 at 08:21, Patrick S. Market wrote:
> > > Steve:
> > > 
> > > I did the touch approach, ran make, and now I get 
> > > 
> > >    Isentropic Highlander =>ar tv $GEMLIB/gemlib.a | grep dftmik
> > >    rw-rw-r--  1001/    10   2228 Mar 29 09:09 2005 dftmik.o
> > > 
> > > However, I get the same error when I try to get a TMIK plot in
> > > GDCNTR:
> > > 
> > >    GEMPAK-GDCNTR>r
> > >    Creating process: gif for queue 2209
> > >    [DG -7]  Input grid TMIK ^970410/0000F000 @900 %PRES cannot be found.
> > > 
> > > I will send the files separately.
> > > 
> > > Pat
> > > 
> > > 
> > > > 
> > > > Patrick,
> > > > 
> > > > Can you send me your makefile and your dftmik.f?
> > > > I'll check the Makefile syntax.
> > > > 
> > > > You can run:
> > > > ar tv $GEMLIB/gemlib.a | grep dftmik
> > > > 
> > > > If that shows that your library does have the dftmik.o routine, then 
> > > > you have
> > > > already compiled it and added it to the library. Make will only compile 
> > > > the
> > > > routine if that is out of date with respect to the source. Editing
> > > > your file or typing "touch dftmik.f" would change the time stamp, so 
> > > > that
> > > > make would have to update the file in the library.
> > > > 
> > > > Steve Chiswell
> > > > 
> > > > On Mon, 28 Mar 2005, Patrick S. Market wrote:
> > > > 
> > > > > Steve:
> > > > >
> > > > > I have, indeed, put dftmik.o into the Makefile.  Typing 'make'
> > > > > in that directory seems to accomplish nothing.  No .o files
> > > > > appear.
> > > > >
> > > > > What shall I try now?
> > > > >
> > > > > Pat
> > > > >
> > > > > >
> > > > > > Patrick,
> > > > > >
> > > > > > The message from "ar" implies that the dftmik.o file has not been
> > > > > > created by compiling the dftmik.f source. Verify that you have added
> > > > > > your dftmik.o routine to the Makefile and that the file succeeds in
> > > > > > compiling.
> > > > > >
> > > > > > Steve Chiswell
> > > > > > Unidata User Support
> > > > > >
> > > > > >
> > > > > > On Tue, 2005-03-22 at 13:39, Patrick S. Market wrote:
> > > > > > > Steve:
> > > > > > >
> > > > > > > As before, I followed those instructions.  I used TMWK as a
> > > > > > > template, since it was a quite similar value that I had hoped to
> > > > > > > generate for the grid.
> > > > > > >
> > > > > > > I have rechecked your instructions, and I did follow them.  I 
> > > > > > > should
> > > > > > > add that even after a make clean/all/install, the command
> > > > > > >
> > > > > > > ar rv $GEMLIB/gemlib.a dftmik.o
> > > > > > >
> > > > > > > returns
> > > > > > >
> > > > > > > ar: cannot open dftmik.o
> > > > > > >         No such file or directory
> > > > > > >
> > > > > > > I will begin again, and try rewriting the tmwk sub from scratch.
> > > > > > >
> > > > > > > Thanks,
> > > > > > > Pat
> > > > > >
> > > > >
> > > > >
> > > > > --
> > > > > ======================================================================
> > > > > Dr. Patrick S. Market
> > > > > Dept. of Soil, Env. & Atmospheric Sci.
> > > > > University of Missouri-Columbia
> > > > > 331 ABNR Hall                                 Phone:  (573) 882 - 1496
> > > > > Columbia, MO 65211 USA        E-mail: address@hidden
> > > > > ======================================================================
> > > > >
> > > > 
> > > 
> > 
>