Mercurial > octave

comparison libinterp/corefcn/data.cc @ 25286:44c0aca121cc stable

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
doc: Clarify docstring for sum (bug #53698). * data.cc: Provide a better explanation of "native" argument and what happens when it is not present.
author Rik <rik@octave.org>
date Thu, 19 Apr 2018 08:39:45 -0700
parents 078b795c5219
children c1e344166c4b e0906c38f18e
comparison
equal deleted inserted replaced
25284:78fb24bdd8bb 25286:44c0aca121cc
2856 Sum of elements along dimension @var{dim}. 2856 Sum of elements along dimension @var{dim}.
2857 2857
2858 If @var{dim} is omitted, it defaults to the first non-singleton dimension. 2858 If @var{dim} is omitted, it defaults to the first non-singleton dimension.
2859 2859
2860 The optional @qcode{"type"} input determines the class of the variable 2860 The optional @qcode{"type"} input determines the class of the variable
2861 used for calculations. If the argument @qcode{"native"} is given, then 2861 used for calculations. By default, operations on floating point inputs (double
2862 the operation is performed in the same type as the original argument, rather 2862 or single) are performed in their native data type, while operations on
2863 than the default double type. 2863 integer, logical, and character data types are performed using doubles. If the
2864 argument @qcode{"native"} is given, then the operation is performed in the same
2865 type as the original argument.
2864 2866
2865 For example: 2867 For example:
2866 2868
2867 @example 2869 @example
2868 @group 2870 @group
2871 sum ([true, true], "native") 2873 sum ([true, true], "native")
2872 @result{} true 2874 @result{} true
2873 @end group 2875 @end group
2874 @end example 2876 @end example
2875 2877
2876 On the contrary, if @qcode{"double"} is given, the sum is performed in 2878 If @qcode{"double"} is given the sum is performed in double precision even for
2877 double precision even for single precision inputs. 2879 single precision inputs.
2878 2880
2879 For double precision inputs, the @qcode{"extra"} option will use a more 2881 For double precision inputs, the @qcode{"extra"} option will use a more
2880 accurate algorithm than straightforward summation. For single precision 2882 accurate algorithm than straightforward summation. For single precision
2881 inputs, @qcode{"extra"} is the same as @qcode{"double"}. Otherwise, 2883 inputs, @qcode{"extra"} is the same as @qcode{"double"}. For all other data
2882 @qcode{"extra"} has no effect. 2884 type @qcode{"extra"} has no effect.
2883 @seealso{cumsum, sumsq, prod} 2885 @seealso{cumsum, sumsq, prod}
2884 @end deftypefn */) 2886 @end deftypefn */)
2885 { 2887 {
2886 int nargin = args.length (); 2888 int nargin = args.length ();
2887 2889