Machine Learning: Using EmbedPy To Apply LASSO Regression

By Samantha Gallagher

The use of kdb+ for machine learning in financial technology and other industries is expanding following the release by KX of the powerful embedPy interface, which allows the kdb+ interpreter to manipulate Python objects, call Python functions, and load Python libraries. Now Python and kdb+ developers can fuse both technologies together, allowing for a seamless application of q’s high-speed analytics and Python’s expansive collection of libraries.

In our latest technical white paper, KX engineer Samantha Gallagher introduces embedPy, covering both a range of basic tutorials as well as a comprehensive solution to a machine-learning project. EmbedPy is available on GitHub to use with kdb+ V3.5+ and Python 3.5 or higher, for macOS or Linux operating systems and Python 3.6 or higher on the Windows operating system. The installation directory also contains a README.txt about embedPy, and an example directory containing thorough examples.

You can read Samantha’s paper on the KX Developer’s site, code.kx.com here.