>
> I think you are stuck having to convert to single precision.
> To get the routines to accept double precision data, the
> ncar graphics low level utilities would have to be rebuilt
> with the above compiler options and they would then not
> work for single precision. However an NCAR graphics
> developer at UCAR might have a trick to use double
> precision arguments that I have not thought of.
We have a double precision version of NCAR Graphics available for Suns
and SGIs. Basically we took the code and ran it through a
float-to-double converter. You won't need to change any of your
graphics calls, but you will need to have your data in double
precision and all reals passed to the routines have to be 8 bytes
(either from C or Fortran codes).
If you are interested in this version and are already licensed for
NCAR Graphics Version 4.0.1, then have your NCAR Graphics site installer
send us email and we'll get you a double precision version.
--Mary
> -------------------------------------------------
> response to Helin Wei <whl@dec200.tea.ac.cn>
>
>
> >I'm trying to make some plots from the double-precison data produced in
> >SUN Ultra2 with f77 -r8 -i4 compiler, but it doesn't work after compiled
> >by ncargf77 -r8 -i4. Here is my question: Can Ncar Graphic produce plots
> >from double-precison data directly?So far,I have to transfer the data to
> >single-precison first then make plots. Any advice?
>
>
This archive was generated by hypermail 2b29 : Wed Jun 28 2000 - 09:45:40 MDT