krr-py

kernel ridge regression utility for simple data files

example usage:

• Simplest usage:

krr.py example.dat

• 2 indep variables columns, custom sigma and alpha:

krr.py -n 2 -s 0.001 -a 0.1 example.dat

• predicted value ranges:

krr.py -r 1:4:10,0:0.3:3 example.dat

• custom range file (observed variables in this file are ignored)

krr.py -f example.dat -n 1 example.dat

dependencies:

• numpy version 1.8.2
• mlpy version 2.2.0

TODO:

• automatic determination of optimal sigma and alpha
• normalization of independent variables so that sigma and alpha remain optimal for all input dimensions
• nicer output formatting