A fork of the original repo from google code (find it here).
Note: If you want to build the tests and samples, just uncomment out the commented lines in wscript
and remove the very last line.
To install
$ git clone https://github.com/mooreryan/redsvd.git
$ cd redsvd
$ make
You'll need a C++ compiler as well as Python to run the waf
build script.
When using sparse matrices and the SVD option, the program spits out some additional files:
- the
US
matrix - the
VS
matrix - the all vs. all cosine dissimilarity matrix of either
US
orVS
, whichever has fewer rows - the
US
vsVS
cosine dissimilarity matrix ifUS
has fewer rows thanVS
, orVS
vsUS
otherwise.
Also, it builds against the Eigen library that is included in this repository to make it an all in one package.