Smile v1.4.0 Release Notes

Release Date: 2017-08-06 // over 6 years ago
  • ๐Ÿ‘€ 1. Add smile-netlib module that leverages native BLAS/LAPACK for matrix computation. See below for the details how to enable it.

    1. Add t-SNE implementation. ๐ŸŽ 3. Improve LLE and Laplacian Eigenmaps performance.
    2. Export DecisionTree and RegressionTree to Graphviz dot file for visualization.
    3. Smile shell is now based on Scala 2.12. ๐Ÿ›  6. Bug fixes.

    โšก๏ธ To enable machine optimized matrix computation, the users should add
    the dependency of smile-netlib:

        <dependency>
          <groupId>com.github.haifengl</groupId>
          <artifactId>smile-netlib</artifactId>
          <version>1.4.0</version>
        </dependency>
    

    and also make their machine-optimised libblas3 (CBLAS) and liblapack3 (Fortran)
    available as shared libraries at runtime.

    OS X

    Apple OS X requires no further setup as it ships with the veclib framework.

    ๐Ÿง Linux

    Generically-tuned ATLAS and OpenBLAS are available with most distributions
    ๐Ÿ“ฆ and must be enabled explicitly using the package-manager. For example,

    • sudo apt-get install libatlas3-base libopenblas-base
    • โšก๏ธ sudo update-alternatives --config libblas.so
    • โšก๏ธ sudo update-alternatives --config libblas.so.3
    • โšก๏ธ sudo update-alternatives --config liblapack.so
    • โšก๏ธ sudo update-alternatives --config liblapack.so.3

    ๐Ÿ— However, these are only generic pre-tuned builds. If you have an Intel MKL licence,
    you could also create symbolic links from libblas.so.3 and liblapack.so.3 to libmkl_rt.so
    or use Debian's alternatives system.

    ๐Ÿ Windows

    ๐Ÿ— The native_system builds expect to find libblas3.dll and liblapack3.dll
    on the %PATH% (or current working directory). Besides vendor-supplied
    implementations, OpenBLAS provide generically tuned binaries, and it
    ๐Ÿ— is possible to build ATLAS.