To install:

1. Use 'make install' to install the module in the system library (or
    use 'make install-user' to install in the user's library). The numpy,
    gvar, and lsqfit Python modules are installed if not already present.

    N.B. Rather than using make and the makefile, one can instead use
    standard python installation tools: run 'pip install .' to install
    to the system library (or 'pip install . --user' to install to
    the user's directory) from within the top directory.

    Install the scipy Python module if you want to use the
    corrfitter.EigenBasis class. Install the matplotlib Python module
    to use CorrFitter.show_plots() for plots of the fits. Neither
    module is not installed automatically.

2. (Optional) Use 'make tests' to run a series of unittests that verify
   that the major components of corrfitter are working. The tests have been
   successful if the output ends with "OK". If 'make' is unavailable
   (non-unix systems), 'python -m unittest discover' can be used to run the
   tests from inside the tests/ directory.

3. (Optional) Look at the examples of corrfitter use in the examples/
   directory; run them all using 'make run-examples'.

4. (Optional) Use 'make clean' to erase work files and directories.


To uninstall:

1. 'make uninstall' (probably) uninstalls most of what was installed. Use
   'make PYTHON=python3 uninstall' to uninstall an installation for a
   different version of Python (python3, in this example).


Problems:

	Contact Peter Lepage: g.p.lepage@cornell.edu


Created by G. Peter Lepage (Cornell University) on 2008-02-12.
Copyright (c) 2008--2018 G. Peter Lepage.
All rights reserved.
