SciPy

SciKits

SciKits (short for SciPy Toolkits) are add-on packages for SciPy, hosted and developed separately and independently from the main SciPy distribution. All SciKits are licensed under OSI-approved licenses.

Packages are packaged as toolkits (instead of in the main, monolithic SciPy distribution) when:

  • The package is deemed too specialized to live in SciPy itself or

  • The package has a GPL (or similar) license which is incompatible with SciPy’s BSD license or

  • The package is meant to be included in SciPy, but development is still in progress.

List of SciKits

An automatically maintained list of SciKits can be found here:

Making SciKits

If you are planning to write a scientific open-source software package for Python, aimed to supplement the existing ones, it may make sense to brand it as a SciKit. The advantage that consistent naming brings is that the package becomes easier to discover, rather than being one amongst the 30000+ Python packages unrelated to research.

The technical step to make this happen is simple: register the project in the Python package index with a name chosen to start with scikit. Examples of existing packages: scikit-learn (machine learning), scikits.sparse (additional sparse matrix routines).

Naturally, you are fully in charge of your own project, and can host it wherever you like. The only requirement we ask is that the project is under an OSI-approved open source license, and is related to science, engineering, or research in some other topic. It can, however, be useful to let people know about it by discussing the new project e.g., on the Scipy mailing lists.

Whilst the recommended license for SciKits projects is the (new) BSD license, SciKits packages are free to choose their own open source license. The license should be officially OSI approved. We, the scipy-developers, will allow packages to contain code with licenses that, in our judgment, comply with the Open Source Definition but have not gone through the approval process. This is to allow us to adopt old code with permissive licenses. The package itself, though, should use a well-known OSI-approved license.

SciKit project structure

An example of a project organization can be found here:

The important part here is the project name passed to the build system, and registering the package in the Python Package Index.

The example package also shows how to use NumPy’s test framework, use Sphinx to write documentation, and link to Fortran code using NumPy’s build framework.

Further reading: