nptsne is a numpy compatible python binary package that offers a number of APIs for fast tSNE calculation and HSNE modelling.
For more detail see the documentation for the current release – 1.1.0
Currently python 3.6, 3.7, and 3.8 are supported on Windows, Mac and Linux using cibuildwheel
Demo software using nptsne
The requirements.txt and the pyproject.toml contain the list of build requirements.
- Add the biovault conan remote (for prebuilt packages):
conan remote add conan-biovault http://cytosplore.lumc.nl:8081/artifactory/api/conan/conan-local
Development build & install using python
pip install -v -e .
This will automatically create a build subdirectory build the bindings and create an .egg-link file in the current python environment.
On Windows a _nptsne.sln file will be present under the build directory
Alternative manual Windows build
- Make a build directory below the HDILib project root. For example: ./_build_release or ./_build_debug (when using conan the source directories are shared but separate build directories should be used for release and debug.)
- In the python environment (with conan and cmake accessible) cd to the build directory and issue the following (for VisualStudio 2017):
cmake .. -G "Visual Studio 15 2017 Win64" -DCMAKE_BUILD_TYPE=Release -DNPTSNE_BUILD_WITH_CONAN=ON
(*Note: this assumes that the build dir is one level down from the project root. The default of NPTSNE_BUILD_WITH_CONAN is OFF*)
- If all goes well Conan will have installed the dependencies in its cache and created the required defines for the Cmake configuration. Open the .sln in VisualStudio and build ALL_BUILD for Release or Debug matching the CMAKE_BUILD_TYPE. On Windows the result of the build are three *.lib files