$prog $version $release_date Copyright (c) $yearrange $author ($author_email) $url INSTALLATION ============ You do not need to install the program in any system directories in order to use it. For testing purposes, I recommend you just unpack the source and run it from there. See the README for more instructions. If you do want to install in system directories, I recommend using setuptools, the de-facto package management system in Python. Setuptools provides a program called easy_install that manages dependencies among packages. You'll have to install wxPython separately, and if you don't have setuptools you can bootstrap it using: http://peak.telecommunity.com/dist/ez_setup.py Once setuptools is installed, you are ready to install with: $ tar xfv peppy-$version.tar.bz2 $ cd peppy-$version $ easy_install -UZ . If you're looking for more info on how to install optional packages, read on... INSTALLING OPTIONAL MODULES =========================== If using the hyperspectral image viewing capabilities of peppy, numpy *must* be installed, and scipy and gdal can optionally be installed. Installing NumPy ---------------- The numerical python library, NumPy (numpy.scipi.org) is installed like a typical python module: by downloading it and running $ python setup.py install You may have to be root to complete the installation. Installing SciPy ---------------- SciPy is a bit more complicated of an install, as it depends on the BLAS and LAPACK libraries. The SciPy website has instructions here: http://www.scipy.org/Installing_SciPy/BuildingGeneral which is, unfortunately, a bit more complicated than just "python setup.py install" but is relatively straight-forward. In general, you can just follow those instructions. However, a problem that plagued me for a while was that the "python setup.py install" stage was picking up libblas and liblapack from the system's /usr/lib directory, and RedHat prior to early 2007 supplied incomplete libraries. I had to replace these system libraries with my just-built libraries, because setup.py seemed to ignore the paths I specified for BLAS and LAPACK as shown in the instructions on the webpage above. Installing GDAL --------------- To take advantage of the variety of image formats that GDAL supports, you can install it as well. Download from www.gdal.org and uncompress in a directory. The setup is a bit more complicated than a typical python module: $ cd gdal-1.4.1 $ ./configure --without-python --with-ngpython $ make $ cd swig/python $ python setup.py install