...

These instructions should be performed as the glastrm user or with appropriate priviledges privileges to write in the GLAST_EXT area.
Note that the scipy unit test will report failures due to missing symbols in the rhel4 blas installed at SLAC.  This is likely always been the case, and apparently cannot be fixed unless we're will to install BLAS and Lapack ourselves.

1. Download Python 2.7.1 from http://www.python.org/ftp/python/2.7.1/Python-2.7.1.tgz and save to /afs/slac/g/glast/ground/GLAST_EXT/srcExtLibs if it doesn't already exist.
2. Change directory to /afs/slac.stanford.edu/g/glast/ground/GLAST_EXT/redhat4-i686-32bit-gcc34/python
3. Ungzip and Untar the file Python-2.7.1.tgz
4. Change directory to Python-2.7.1
5. edit setup.py, in detect_modules() change the lines from:

...

1. Execute ./configure --prefix /afs/slac.stanford.edu/g/glast/ground/GLAST_EXT/redhat4-i686-32bit-gcc34/python/2.7.1/gcc34
2. Execute make
3. Execute make install
4. Leave and erase the current directory (Python-2.7.1)

...

redhat4-x86-64bit-gcc34

Unable to install numpy 1.5.1 successfully on rhel4-64, due to a known bug with 64 bit and gcc3.  The numpy unit test fails.  Fix is to move to gcc4, which we cannot do.  For now remaining on python 2.6.5.  In either case, the scipy unit tests results some errors which we seemingly have always been dealing with.  The bug report suggests doing a new install of BLAS, but according to unix-admins that is not possible.

redhat5-i686-32bit-gcc41

Same directions as redhat4-i686-32bit-gcc34.  The fortran compiler to use for numpy and scipy install is gfortran so --fcompiler=gnu95.  There should be no errors in the numpy or scipy unit tests.

redhat5-x86_64-64bit-gcc41

Same directions as redhat4-i686-32bit-gcc34.  The fortran compiler to use for numpy and scipy install is gfortran so --fcompiler=gnu95.  There should be no errors in the numpy or scipy unit tests.

snowleopard-gcc42

pip and easy_install have no connectivity to download on bldmac02, so performing all module installations by hand.

Skipping matplotlib (some missing system libraries would have to be installed), pywcsgrid2 (needs matplotlib), APLpy (needs matplotlib), and SciPy (due to missing fortran compiler)

Additional Python Modules

These modules should additionally be installed into python after updating PATH to point to the new python binary.

distribute 0.6.16 for use with pip

1. curl -O http://python-distribute.org/distribute_setup.py
2. python distribute_setup.py

pip 1.0.1

easy_install replacement that should also work with python 3.0

1. curl -O https://github.com/pypa/pip/raw/master/contrib/get-pip.py
2. python get-pip.py

nose 1.0.0

unit test framework necessary for running numpy and scipy unit tests

1. pip install nose

NumPy - 1.5.1

Still installing by hand to include --fcompiler flag to force correct fortran compiler to be used.  This is dictated by whatever blas and lapack are linked against.  For more information, see the numpy INSTALL.txt file.

1. Download numpy-1.5.1.tar.gz from http://sourceforge.net/project/showfiles.php?group_id=1369&package_id=175103&release_id=628858
2. Change directory to /afs/slac.stanford.edu/g/glast/ground/GLAST_EXT/redhat4-i686-32bit-gcc34/python
3. Extract numpy-1.5.1.tar.gz
4. Change directory to numpy-1.5.1
5. python setup.py build --fcompiler=gnu (for g77) or gnu95 for gfortran.
   It seems the rhel4s have blas and lapack linked against g77, while rhel5 uses gfortran
6. python setup.py install
7. Run the test routine by starting up python and doing:
   import numpy
   numpy.test()
8. Leave and erase the current directory (numpy-1.5.1)
   

PyFITS - 2.4.0

1. pip install pyfits

matplotlib 1.0.0

Had to provide explicit web link to version we wanted to install, otherwise the default seems to be 0.91

1. pip install http://sourceforge.net/projects/matplotlib/files/matplotlib/matplotlib-1.0/matplotlib-1.0.0.tar.gz

pywcs-1.10-4.7

1. pip install pywcs

APLpy-0.9.5

1. pip install APLpy

4Suite XML - 1.0.2

This module was not recognized by pip.  Using easy_install instead.

1. easy_install 4Suite-XML

SciPy - 0.9.0

Continuing to install by hand to provide compiler flag to force correct fortran compiler, depending on what blas and lapack are linked against (try ldd /usr/lib/libblas.so).  See SciPy's INSTALL.txt file for more information.

1. Download scipy-0.9.0

These modules should additionally be installed into python after updating PATH to point to the new python binary.

distribute 0.6.16 for use with pip

1. curl -O http://python-distribute.org/distribute_setup.py
2. python distribute_setup.py

pip 1.0.1

easy_install replacement that should also work with python 3.0

1. curl -O https://github.com/pypa/pip/raw/master/contrib/get-pip.py
2. python get-pip.py

nose 1.0.0

unit test framework necessary for running numpy and scipy unit tests

1. pip install nose

NumPy - 1.5.1

Still installing by hand to include --fcompiler flag to force correct fortran compiler to be used.  This is dictated by whatever blas and lapack are linked against.  For more information, see the numpy INSTALL.txt file.

1. Download numpy-1.5.1.tar.gz from http://sourceforge.net/project/showfiles.php?group_id=1369&package_id=175103&release_id=628858
2. Change directory to /afs/slac.stanford.edu/g/glast/ground/GLAST_EXT/redhat4-i686-32bit-gcc34/python
3. Extract numpy-1.5.1.tar.gz
4. Change directory to numpy-1.5.1
5. python setup.py build --fcompiler=gnu (for g77) or gnu95 for gfortran.
   It seems the rhel4s have blas and lapack linked against g77, while rhel5 uses gfortran
6. python setup.py install
7. Run the test routine by starting up python and doing:
   import numpy
   numpy.test()
8. Leave and erase the current directory (numpy-1.5.1)
   

PyFITS - 2.4.0

1. pip install pyfits

matplotlib 1.0.0

Had to provide explicit web link to version we wanted to install, otherwise the default seems to be 0.91

1. pip install http://sourceforge.net/projects/matplotlib/files/matplotlib/matplotlib-1.0/matplotlib-1.0.0.tar.gz^{Image Removed}

pywcs-1.10-4.7

1. pip install pywcs

APLpy-0.9.5

1. pip install APLpy

4Suite XML - 1.0.2

This module was not recognized by pip.  Using easy_install instead.

1. easy_install 4Suite-XML

Scipy - 0.9.0

1. Download ipython-0.10.tar.gz from http://www.scipy.org/ and save to /afs/slac.stanford.edu/g/glast/ground/GLAST_EXT/redhat4-i686-32bit-gcc34/python if it doesn't already exist.
2. Change directory to /afs/slac.stanford.edu/g/glast/ground/GLAST_EXT/redhat4-i686-32bit-gcc34/python
3. Extract scipy-0.7.2.tar.gz
4. Change directory to scipy-0.7.2
5. Execute /afs/slac.stanford.edu/g/glast/ground/GLAST_EXT/redhat4-i686-32bit-gcc34/python/2.6.5/gcc34/bin/python setup.py install
6. Leave and erase scipy-0.7.2

iPython - 0.10.2

1. pip install ipython

Scipy - 0.7.2

1. Download ipython-0.10.tar.gz from http://www.scipy.org/ and save to /afs/slac.stanford.edu/g/glast/ground/GLAST_EXT/redhat4-i686-32bit-gcc34/python if it doesn't already exist.
2. Change directory to /afs/slac.stanford.edu/g/glast/ground/GLAST_EXT/redhat4-i686-32bit-gcc34/python
3. Extract scipy-0.79.20.tar.gz
4. Change directory to scipy-0.7.2
5. Execute /afs/slac.stanford.edu/g/glast/ground/GLAST_EXT/redhat4-i686-32bit-gcc34/python/2.6.5/gcc34/bin/python setup.py install
7. Change directory to scipy-0.9.0
8. Execute python setup.py config_fc --fcompiler=gnu install (gnu for g77 or gnu95 for gfortran) 
9. Leave and erase scipy-0.9.0
10. Test by starting up python and doing:
    import scipy
    scipy.test()

iPython - 0.10.2

1. pip install ipython
2. Run the iptest located in python's bin directory.

pywcsgrid2

1. Download from https://github.com/leejjoon/pywcsgrid2/downloads^{Image Added}
2. Extract
3. cd
4. python setup.py installLeave and erase scipy-0.7.2

2.5.1

These are instructions for building Python 2.5.1 for the various OSes supported.

...