Releases: tassia/AppRecommender
Releases · tassia/AppRecommender
v0.7.5
v0.7.4
Bump setup.py version
v0.7.3
v0.7.2
- Fix bug on training without sufficient packages
- Add integration with apt
- Add update command that runs both init and train commands
v0.7.1
- Fix bug when there is no packages for machine learning training process
v0.7
- Create new strategy that allows a package to be used as an reference.
This new strategy that a package and together with the other user's packages, it will then recommend packages based on both the reference package and the packages the user already has.
- Update how --init remove directories
Now, --init do not erase the base directory that hold the AppRecommender files. It erases just the necessary directories inside this base directory, if these directories already exists
v0.6.4
- Improve AppRecommender command line parsing
- Allow AppRecommender to read from environment variable to set its install path
v0.6.3
- Change how log creating is handled
- Allow AppRecommender to install files on /var/lib
- Fix modules that are installed on the user system
- Change how AppRecommender is executed
v0.6.2.3
- Change shell option for subprocess.Popen on collect_user_data.py
- Remove commented code from recommender.py
v0.6.2.2
- Bump version on setup.py
- Add missing dependency -python-fuzzywuzzy on README
- Remove unused code from user.py