Fisher matrices encode the cosmological constraints (exepcted or actual) from a given experiment (e.g., weak lensing with JDEM). Fisher.py allows you to combine constraints from multiple experiments (e.g., weak lensing + supernovae) and add priors (e.g., a flat universe) simply and easily. Calculate parameter uncertainties and plot confidence ellipses as shown below. Fisher matrix expectations for several experiments are included as calculated by myself (time delays) and the Dark Energy Task Force (WL/SN/BAO/CL/CMB), or provide your own.
For an introduction to Fisher matrices and confidence contours, see
Fisher Matrices and Confidence Ellipses: A Quick Start Guide and Software (Coe 2009).
This was spawned from my work on
Cosmological Constraints from Gravitational Lens Time Delays (Coe & Moustakas 2009).
Extract by running "tar -xvf Fisher2.tar.gz" (or double-clicking on a Mac).
Set the following environmental variables:
setenv NUMERIX numpy
setenv FISHERPATH /Users/coe/Fisher (or wherever you put the software)
setenv PYTHONPATH {$PYTHONPATH}:{$FISHERPATH}
Test programs (try running these):
python simple.py python plotexps.py
Mac installation (Leopard OS X 10.5):
Python should already be installed.
For Numpy & Matplotlib, I recommend the
Scipy Superpack.
After downloading, install by running "sh superpack_2009.03.28.sh".
Then to enable arrow key commands, remove / rename the file
"/Library/Python/2.5/site-packages/pyreadline-1.6.svn.dev_r3066-py2.5.egg".