Get the fortran subroutine here

Get the sample data: va_djf.nc and ps.nc 

The easiest way to wrap this using f2py is to run the following comand

f2py -c -m streamfunction sf.f90

You should get something similar to the following on your screen:

running build
running config_cc
unifing config_cc, config, build_clib, build_ext, build commands --compiler options
running config_fc
unifing config_fc, config, build_clib, build_ext, build commands --fcompiler options
running build_src
building extension "streamfunction" sources
f2py options: []
f2py:> /tmp/tmpwrCzXz/src.linux-i686-2.5/streamfunctionmodule.c
creating /tmp/tmpwrCzXz
creating /tmp/tmpwrCzXz/src.linux-i686-2.5
Reading fortran codes...
        Reading file 'sf.f90' (format:free)
Post-processing...
        Block: streamfunction
                        Block: ccmp_zm_mspi
Post-processing (stage 2)...
Building modules...
        Building module "streamfunction"...
                Constructing wrapper function "ccmp_zm_mspi"...
                  zm_mpsi = ccmp_zm_mspi(v,lat,p,ps,msg,[nlon,nlat,nlev])
        Wrote C/API module "streamfunction" to file "/tmp/tmpwrCzXz/src.linux-i686-2.5/streamfunctionmodule.c"
  adding '/tmp/tmpwrCzXz/src.linux-i686-2.5/fortranobject.c' to sources.
  adding '/tmp/tmpwrCzXz/src.linux-i686-2.5' to include_dirs.
copying /lgm/cdat/latest/lib/python2.5/site-packages/numpy/f2py/src/fortranobject.c -> /tmp/tmpwrCzXz/src.linux-i686-2.5
copying /lgm/cdat/latest/lib/python2.5/site-packages/numpy/f2py/src/fortranobject.h -> /tmp/tmpwrCzXz/src.linux-i686-2.5
running build_ext
customize UnixCCompiler
customize UnixCCompiler using build_ext
customize GnuFCompiler
Could not locate executable g77
Could not locate executable f77
customize IntelFCompiler
Could not locate executable ifort
Could not locate executable ifc
customize LaheyFCompiler
Could not locate executable lf95
customize PGroupFCompiler
Could not locate executable pgf90
Could not locate executable pgf77
customize AbsoftFCompiler
Could not locate executable f90
customize NAGFCompiler
Could not locate executable f95
customize VastFCompiler
customize GnuFCompiler
customize CompaqFCompiler
Could not locate executable fort
customize IntelItaniumFCompiler
Could not locate executable efort
Could not locate executable efc
customize IntelEM64TFCompiler
customize Gnu95FCompiler
Found executable /usr/local/bin/gfortran
customize Gnu95FCompiler
customize Gnu95FCompiler using build_ext
building 'streamfunction' extension
compiling C sources
C compiler: gcc -fno-strict-aliasing -DNDEBUG -O -fPIC

creating /tmp/tmpwrCzXz/tmp
creating /tmp/tmpwrCzXz/tmp/tmpwrCzXz
creating /tmp/tmpwrCzXz/tmp/tmpwrCzXz/src.linux-i686-2.5
compile options: '-I/tmp/tmpwrCzXz/src.linux-i686-2.5 -I/lgm/cdat/latest/lib/python2.5/site-packages/numpy/core/include -I/lgm/cdat/latest/include/python2.5 -c'
gcc: /tmp/tmpwrCzXz/src.linux-i686-2.5/fortranobject.c
gcc: /tmp/tmpwrCzXz/src.linux-i686-2.5/streamfunctionmodule.c
compiling Fortran sources
Fortran f77 compiler: /usr/local/bin/gfortran -Wall -ffixed-form -fno-second-underscore -fPIC -O3 -funroll-loops -march=pentium4 -mmmx -msse2 -msse -fomit-frame-pointer -malign-double
Fortran f90 compiler: /usr/local/bin/gfortran -Wall -fno-second-underscore -fPIC -O3 -funroll-loops -march=pentium4 -mmmx -msse2 -msse -fomit-frame-pointer -malign-double
Fortran fix compiler: /usr/local/bin/gfortran -Wall -ffixed-form -fno-second-underscore -Wall -fno-second-underscore -fPIC -O3 -funroll-loops -march=pentium4 -mmmx -msse2 -msse -fomit-frame-pointer -malign-double
compile options: '-I/tmp/tmpwrCzXz/src.linux-i686-2.5 -I/lgm/cdat/latest/lib/python2.5/site-packages/numpy/core/include -I/lgm/cdat/latest/include/python2.5 -c'
gfortran:f90: sf.f90
sf.f90:107.56:

  INTEGER                                           :: k
                                                       1
Warning: Unused variable k declared at (1)
sf.f90: In function ‘ccmp_zm_mspi’:
sf.f90:106: warning: ‘flag’ may be used uninitialized in this function
/usr/local/bin/gfortran -Wall -Wall -shared /tmp/tmpwrCzXz/tmp/tmpwrCzXz/src.linux-i686-2.5/streamfunctionmodule.o /tmp/tmpwrCzXz/tmp/tmpwrCzXz/src.linux-i686-2.5/fortranobject.o /tmp/tmpwrCzXz/sf.o -lgfortran -o ./streamfunction.so
Removing build directory /tmp/tmpwrCzXz

Most importantly you should now have a file called:

This file can be loaded in Python

Let's take a look at it, from the directory where the file resides run python, import the file/module

python
Python 2.5.1 (r251:54863, Jan  2 2008, 11:40:37)
[GCC 4.2.1] on linux2
Type "help", "copyright", "credits" or "license" for more information.
Loaded 5.0 settings


>>> import streamfunction

We can look at the module contents:

>>> dir(streamfunction)
['__doc__', '__file__', '__name__', '__version__', 'ccmp_zm_mspi']

Not surprisingly it only contains the ccmp_zm_mspi subroutine

Let's take a look at it, f2py nicely documented its call for us:

>>> print streamfunction.ccmp_zm_mspi.__doc__
ccmp_zm_mspi - Function signature:
  zm_mpsi = ccmp_zm_mspi(v,lat,p,ps,msg,[nlon,nlat,nlev])
Required arguments:
  v : input rank-3 array('f') with bounds (nlon,nlat,nlev)
  lat : input rank-1 array('f') with bounds (nlat)
  p : input rank-1 array('f') with bounds (nlev)
  ps : input rank-2 array('f') with bounds (nlon,nlat)
  msg : input float
Optional arguments:
  nlon := shape(v,0) input int
  nlat := shape(v,1) input int
  nlev := shape(v,2) input int
Return objects:
  zm_mpsi : rank-2 array('f') with bounds (nlat,nlev)

Ok, at this stage it is important to note a few things.

1- f2py separated input and output arguments, the later ones are now returned (zm_mpsi)

2- f2py detected that some input arguments can be deducted from the other input arguments, in this case the arrays dimensions

   This is important for 2 reason:

       a- You do not need to pass these, f2py will figure it out on its own

       b- If you decide to pass them anyway, they have been shifted to the end of the call, remember in the fortran call nlon, nlat and nlev were the first arguments to pass, now they do not need to be passed but must be passed last.
 

OK let's use our sample dataset and call this function

import streamfunction
import cdms2
# Open data file 1
f=cdms2.open('va_djf.nc')
# reads in data, makes sure order is right and levels are top to bottom
v=f('v',level=(0,1E10),order='xyz') # right order, p starts at 0 (top)
f.close()
# now get the latitude and level axes
lat=v.getLatitude()
p=v.getLevel()
# does the same with second file
fp=cdms2.open('ps.nc')
ps=fp('ps',order='xy') # right order
fp.close()
msg=1.e20

zm_mpsi=streamfunction.ccmp_zm_mspi(v.filled(),lat[:],p[:],ps.filled(),msg)

Let's point out a few important things

1- When reading the data we made sure they'd be ordered according to what the fortran wants (lon,lat,lev)

2-

  a- When passing or variable to the fortran we convert them to numpy object, indeed only understand pure array, it does not know about masked values or axis or attributes.

  b- Axes object had to be converted as well to pure array , that's why we used [:] when passing them, it returns their values only as an array

3- When converting MV/MA to numpy pure arrays we used the .filled() attribute, you can supply a value to this function, this value will be used to replace missing values, the default is 0, but sometime your fortran subroutine might excpect another value (say -999), in this case you'd pass v.filled(-999.)

Here you are, it's all done, in the next step we'll show how to write a little python function to pre/post process your data automatically.