8.1 The DyND Library (invited)

Wednesday, 9 January 2013: 10:30 AM
Room 12B (Austin Convention Center)
Mark Wiebe, Continuum Analytics, Austin, TX

DyND is a C++ library which extends the dynamic data typing model of NumPy to allow the use of NumPy's array programming style on data which has a more general structure than fits in NumPy arrays. Expressions are evaluated in a lazy fashion, and dynamic code generation is used to efficiently adapt a variety of input code styles to the data model. It features a syntax natural syntax from C++, similar in style to boost multiarray but more dynamic in nature, and full exposure to Python with automatic NumPy interoperability where the data can be represented in both systems.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner