BIG DATA

New Python wrappers for the IMSL C library deliver powerful prototype to production numerical application development: Visual Numerics, a 37-year producer of advanced numerical analysis and visualization software, announced the immediate availability of PyIMSL, a pre-built, fully-tested and documented collection of Python wrappers to the mathematical and statistical algorithms in the IMSL C Numerical Library. With PyIMSL, developers can use Python and the IMSL C Library for rapid prototyping of numerical applications. For production applications, they can then directly deploy the Python application into production or if they choose to rewrite the application in C/C++, use the same mathematical and statistical algorithms in both development environments. PyIMSL is unique in that developers get consistency between the algorithms used for prototyping as used for production, reducing the cost and time-to-market to move numerical applications from prototype to production. Visual Numerics has developed Python wrappers for all IMSL C Library numerical analysis functionality. Developers can now write applications in Python and have PyIMSL wrappers call code in IMSL C Library DLLs or shared libraries. The PyIMSL API closely matches the IMSL C Library API, so any applications that are taken from a Python prototype to a C/C++ production environment will make use of the IMSL C Library mathematics and statistics functions by simply calling the same API. With no differences in prototype and production algorithms, developers can create applications quicker and with fewer differences that need to be resolved. "As a long-time technical advisor to Visual Numerics, I’m pleased to see the company extending the IMSL C Library to Python developers,” said Dr. Gary McClelland, Professor of Psychology at the University of Colorado. “It gives their commercial customers an outstanding option to quickly prototype in Python and then write production applications in C using the same algorithms. For the education community, making the IMSL Libraries accessible to Python developers is important given the growing popularity of the language. Python is very popular on campus these days, but good numerical packages are difficult to find." Python is a popular open source language and growing in popularity. The Tiobe Programming Community Index for August 2008 ranks Python #6, an increase by two spots since the August 2007 rankings. PyIMSL gives developers the option to develop in Python while still leveraging the comprehensive mathematical and statistical algorithms available in the IMSL C Library. Key benefits of PyIMSL include the following:

• Reduced cost and time-to-market to move numerical applications from prototype to production. Because PyIMSL wraps IMSL C Library algorithms, an application written initially in PyIMSL and then ported to C/C++ will leverage the same algorithms and deliver the same results.
• Numerical library accessibility. With PyIMSL, Visual Numerics continues to deliver the industry’s most accessible library of math and statistical algorithms. Developers can now access IMSL Library functionality from Python in addition to having native C, Java, .NET, and Fortran options.
• Advanced numerical analysis for Python. Open source libraries have limited analytics in Python, and current Python tools lack comprehensive and supported analytical libraries. PyIMSL delivers industry-proven and commercially seasoned IMSL C Library algorithms in the Python language.

"We are committed to offering our customers the highest quality, most sophisticated numerical analysis solutions for mainstream platforms," said Sean FitzGerald, Vice President of Technology and Consulting of Visual Numerics. "Visual Numerics is proud to extend the IMSL Libraries to Python and bring such strong analytics to Python developers." Availability PyIMSL is available for immediate use. PyIMSL is available at no charge to current IMSL C Library customers and developers must have the IMSL C Library V5.5 or higher installed to use PyIMSL. Please contact Visual Numerics for details.