Microsoft is now preparing its flagship RDBMS product for machine-learning functions with its integration of the Python programming language. The latest Community Technology Preview 2 (CTP2) release of SQL Server 2017 features Python code alongside additional support for R, first introduced with SQL Server 2016.
New SQL Server Features
Microsoft detailed SQL Server 2017’s upcoming features at length in a blog post, which included several advantages offered by Python, such as eliminating data movement problems, easy deployment of Python models through embedding in a T-SQL script, and vast extensibility to build AI applications on large SQL Server data.
Python’s inclusion in SQL Server owes in large part to its reputation as a leading data science tool, similar to the R programming language. DBA Services agrees that Python’s powerful scripting and noted readability makes it a tool of choice for many developers, IT administrators, data scientists, and analysts. While Microsoft’s acquisition of Revolution Analytics in 2015 signified its strong support for R as the go-to data science-oriented language, its inclusion of Python clarifies its desire to go where developers are.
The Power of Python
Introducing Python/R functionality with SQL Server data lets developers use the new interface for extensive machine learning applications at will on text, images, and other unstructured data. This makes for increased throughput by implementing the resulting analytics in-database as stored procedures.
For early adopters, having an existing Python installation is not required. SQL Server 2017’s set-up process prompts a download and installation of its standard CPython 3.5 interpreter (also available at the Python.org website) once the user chooses Python as one of its SQL Server’s features. Users are also free to install their Python packages to run on SQL Server, like Cognitive Toolkit and TensorFlow for modern deep learning applications.
The inclusion of Python with SQL Server’s already potent interface using R opens up exciting possibilities. The AI component that comes with Python packages, combined with large SQL Server data has the potential to build stronger and more intelligent database applications.