Difference between revisions of "cpp/numeric/valarray"
(mention the (lost) competition with expression templates.) |
m (typo) |
||
| Line 9: | Line 9: | ||
===ONotes=== | ===ONotes=== | ||
| − | {{cpp|std::valarray}} and helper classes are defined to be free of certain forms of aliasing, thus allowing operations on these classes to be optimized similar to the effect of the keyword {{tt|restrict}} in the C programming language. In addition, functions and operators that take {{tt|valarray}} arguments are allowed to return special objects which enable vectorization: an expression such as {{cpp|1=v1 = a*v2 + v3;}} would be compiled as a single loop that executes {{cpp|1= | + | {{cpp|std::valarray}} and helper classes are defined to be free of certain forms of aliasing, thus allowing operations on these classes to be optimized similar to the effect of the keyword {{tt|restrict}} in the C programming language. In addition, functions and operators that take {{tt|valarray}} arguments are allowed to return special objects which enable vectorization: an expression such as {{cpp|1=v1 = a*v2 + v3;}} would be compiled as a single loop that executes {{cpp|1=v1[i] = a*v2[i] + v3[i];}} avoiding any temporaries or multiple passes. However, [[enwiki::Expression templates|expression templates]] make the same optimization technique available for any C++ container, and the majority of numeric libraries prefer expression templates to valarrays. A rare exception is [http://software.intel.com/sites/products/documentation/studio/composer/en-us/2011/compiler_c/cref_cls/common/cppref_valarray_intro.htm Intel Parallel Studio ] |
===Member types=== | ===Member types=== | ||
Revision as of 11:13, 10 January 2012
Template:cpp/numeric/valarray/sidebar
| Defined in header <valarray>
|
||
| template< class T > class valarray; |
||
Template:cpp is the class for representing and manipulating arrays of values. It supports element-wise mathematical operations and various forms of generalized subscript operators, slicing and indirect access.
Contents |
ONotes
Template:cpp and helper classes are defined to be free of certain forms of aliasing, thus allowing operations on these classes to be optimized similar to the effect of the keyword restrict in the C programming language. In addition, functions and operators that take valarray arguments are allowed to return special objects which enable vectorization: an expression such as Template:cpp would be compiled as a single loop that executes Template:cpp avoiding any temporaries or multiple passes. However, expression templates make the same optimization technique available for any C++ container, and the majority of numeric libraries prefer expression templates to valarrays. A rare exception is Intel Parallel Studio
Member types
Template:tdcl list begin Template:tdcl list hitem Template:tdcl list item Template:tdcl list end
