Releases: IBMStreams/streamsx.sparkMLLib
SparkMLLib Toolkit v1.3.3 Release
SparkMLLib Toolkit v1.3.2 Release
New:
- Third-party lib updated to resolve security vulnerabilities (jackson-databind: 2.6.7.3)
Checksums for streamsx.sparkMLLib-1.3.2-434abb3-20200617-1240.tgz
MD5: 7ba93e5bf0fcfca843a7554a51d7563b
sha1: baf87bbbb70c5776c207460e4ee4ccbf6ea8fac2
SparkMLLib Toolkit v1.3.1 Release
New:
- Update internationalization messages
Checksum streamsx.sparkMLLib-1.3.1-a94b360-20200214-1649.tgz
MD5: 8596fa4465b043ae0d5efed4c88902bc
SHA1: a40d8855ed0f3fc619ebf9f9f8108abfb4d03270
SparkMLLib Toolkit v1.3.0 Release
New:
- The toolkit not longer depend on an installation of Apache Spark and does not need a SPARK_HOME environment variable,
- Correct streams studio classpath settings in toolkit project and sample
- Use studio settings in sample makefile if build from studio
- Update description
- Remove compiler warnings
- Describe spark master parameter
- Add framework tests
- Add test and release targets to main build.xml
- New parameter paraneter getProbabilities in operator SparkNaiveBayes
streamsx.sparkMLLib-1.3.0-b0c3923-20191121-1610.tgz MD5: a3e25e1e9893f6cb812e4ea4cc180c16
streamsx.sparkMLLib-1.3.0-b0c3923-20191121-1610.tgz sha1: 39bf40e9c6a35918f3f28af75194483981a251c0
SparkMLLib Toolkit v1.2.0 Release
New:
- Use of actual stark version 2.4.0
SparkMLLib Toolkit v1.1.1 Release
-
Some path changes for build artifacts
-
Internationalization for languages de_DE, es_ES, fr_FR, it_IT, ja_JP, ko_KR, pt_BR, ru_RU, zh_CN, zh_TW
SparkMLLib Toolkit v1.0.0 Release
This release includes operators that support loading and scoring against a variety of Spark MLLib algorithms including:
- Classification
- Linear SVM
- Naive Bayes
- Clustering
- KMeans
- Collaborative Filtering
- Regression
- Isotonic
- Linear
- Logistic
- Tree
- Decision Tree
- Gradient Boosted Trees
- Random Forest
v0.8.0.0 Pre-release
A number of changes were made:
- Operator namespaces were refactored as per issue #3
- Added ability to reload spark models using a control port as per issue #5
- A number of new models were added as per issue #4:
- classification - SparkLinearSVM
- clustering - SparkClusteringKMeans
- regression - SparkIsotonicRegression, SparkLogisticRegression
v0.7.0.0 Pre-release
This pre-release includes operators that support loading some of Apache Spark MLlib algorithms including:
- Collaborative filtering
- Decision tree
- Tree ensembles - Random Forest, Gradient-Boosted Trees
- Linear regression
- Naive Bayes