Spar: a set of extensions to Java for scientific computation

In this paper we present a set of language extensions that improve the expressiveness and performan e of Java for s ienti omputation. First of all, the language extensions allow the manipulation of multi-dimensional arrays to be expressed more naturally, and to be implemented more efiently. Furthermore, data-parallel programming is supported, allowing eÆ ient parallelization of a large lass of operations on arrays. We also provide language extensions to onstru t spe ialized array representations, su h as symmetri , blo k, and sparse matri es. These extensions are: tuples, parameterized types, array subs ript overloading, and the inline modi er. These extensions are not only useful to onstru t spe ial array representations, but are also useful in their own right. Finally, we add omplex numbers as a primitive type to the language. We evaluate our language extensions using performan e results. We also ompare relevant ode fragments of our extended language with standard Java implementations and language extensions proposed by others. General Terms Language Design, S ienti Computation