Revision 0f9ce693bda1e6591173af4de60a2cfa610f4957 authored by Marie Roald on 06 June 2021, 08:48:42 UTC, committed by Marie Roald on 06 June 2021, 08:48:42 UTC
Only making the constrained factor matrices non-negative seemed
to make the fitting procedure less stable (higher likelihood of
all-zero components).
1 parent 5f08046
History
File Mode Size
.github
conda
doc
examples
tensorly
.coveragerc -rw-r--r-- 72 bytes
.gitignore -rw-r--r-- 905 bytes
.travis.yml -rw-r--r-- 2.6 KB
AUTHORS.rst -rw-r--r-- 272 bytes
CHANGELOG.md -rw-r--r-- 229 bytes
CONTRIBUTING.rst -rw-r--r-- 275 bytes
LICENSE.txt -rw-r--r-- 1.5 KB
Makefile -rw-r--r-- 607 bytes
README.rst -rw-r--r-- 7.4 KB
deploy_docs.sh -rw-r--r-- 1.3 KB
requirements.txt -rw-r--r-- 35 bytes
setup.cfg -rw-r--r-- 508 bytes
setup.py -rwxr-xr-x 1.9 KB

README.rst

back to top