Release Announcement: Jubatus 1.0.0

[motoki yuhara]

Hi Jubatus guys, 

We are glad to announce that Jubatus 1.0.0 is now available. 

In this release we have a plenty of new features and improvement.

Note that some changes are incompatible to older Jubatus versions.

- A new feature, Embdded Jubatus Python. This feature makes it possible to use Jubatus Core functions from python codes without launching Jubatus servers.

- An image feature extraction plugin. OpenCV based image feature extraction is now officialy supported.

- A new clustering algorithm, DBSCAN. It can cluster points without specifying a cluster size.

- Official image in Dockerhub.  Jubatus image is now available here and you can easily use Jubatus by `docker pull jubatus/jubatus`.

- Refreshing our Official Website.  You can now enjoy our stylish website.

Note

- jubadump command is now unavailable for Classifier and Regression models. This bug will be fixed in next release.

- For CentOS users who already use jubadump, yum update is now unavailable. Please uninstall jubadump and update jubatus using following commands

  - yum remove jubadump

  - yum update jubatus

  - yum install jubatus

The full changelogs are as follows:

===================================

Jubatus(Core)

-------------

* New Machine Learning Algorithms

  * clustering: add DBSCAN algorithm (#330, #332)

  * regression: add perceptron, PA, PA2, CW, AROW, NHERD, NN, cosine and euclidean (#327, #328)

* Improvements

  * clustering: improve APIs to make each Datum distinguishable by user-specified ID (#317, #326)

  * clustering: improve config format and parameter range checking (#314, #316, #325, #334, #335)

  * graph: add method to return the largest ID (#333)

  * classifier and regression: Rename diff_size to num_features_diff status key (#336)

  * Improve binary feature key names to include the rule name (#329)

  * regression: remove unused argument from driver constructor (#319, #320)

  * Remove unused revert function (#315, #331)

* Bug fixes

  * recommender: fix rows not cleared occasionally in inverted_index and inverted_index_euclid (#318)

Jubatus(Server)

---------------

* New Features

  * Add image extraction plug-in (#843, #1150)

  * Add configuration validation feature (#973, #1144)

* Improvements

  * clustering: change APIs to accept ID-based datum (#1155)

  * Update Dockerfile to use Ubuntu 14.04 (#1146)

  * Improve --daemon option of server and proxy to detach from console (#1072, #1148)

  * classifier: models are now compatible in standalone and distributed mode (#469, #1153)

  * Change the default log level to INFO instead of DEBUG (#1152)

  * Update example configurations (#1165)

  * Update manpages (#1167)

  * Update packaging tools (#1157)

* Bug fixes

  * graph: fix IDs being reused when loading model file to other server instance in standalone mode (#1149, #1156)

We hope you enjoy this release! 

Yuhara Motoki @ Jubatus Team