[Numpy-discussion] FORTRAN compiler detection

Robert Kern robert.kern at gmail.com
Tue Feb 26 14:20:26 EST 2008


On Tue, Feb 26, 2008 at 12:46 PM, Christopher Hanley <chanley at stsci.edu> wrote:
> Robert Kern wrote:
>  > On Tue, Feb 26, 2008 at 12:09 PM, Christopher Hanley <chanley at stsci.edu> wrote:
>  >> Robert Kern wrote:
>  >>  > On Tue, Feb 26, 2008 at 11:50 AM, Christopher Hanley <chanley at stsci.edu> wrote:
>  >>  >> Greetings,
>  >>  >>
>  >>  >>  I was wondering if within the last 8 - 10 weeks anyone has made changes
>  >>  >>  to the way FORTRAN compilers are detected.  In the past I was able to
>  >>  >>  specify which compiler was used by the F77 system variable.  However, I
>  >>  >>  am now having a f90 compiler that exists on my Solaris system detected
>  >>  >>  regardless of the F77 value.  Even unsetting the F77 variable leads to
>  >>  >>  the use of the f90 compiler.
>  >>  >>
>  >>  >>  I was going to look though the distutil change logs but I was hoping
>  >>  >>  that someone might remember changing something off the top of their heads.
>  >>  >
>  >>  > Which FORTRAN compilers do you have installed? What --fcompiler flag
>  >>  > are you using?
>  >>  >
>  >>  Hi Robert,
>  >>
>  >>  Thank you for your help.
>  >>
>  >>  I am not using the --fcompiler flag.
>  >
>  > Okay, use the --fcompiler flag. That is the way to tell numpy to use a
>  > particular compiler if you have multiple ones installed.
>  >
>
>  What do you do if you have FORTRAN compilers installed but don't want to
>  use any of the compilers to build numpy?

Unless if you are linking to a FORTRAN LAPACK or BLAS, no FORTRAN
compiler should be used to build numpy. If you find that one is being
used, please show us the full output log.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco



More information about the NumPy-Discussion mailing list