ssumkbn2.js (2668B)
1 /** 2 * @license Apache-2.0 3 * 4 * Copyright (c) 2020 The Stdlib Authors. 5 * 6 * Licensed under the Apache License, Version 2.0 (the "License"); 7 * you may not use this file except in compliance with the License. 8 * You may obtain a copy of the License at 9 * 10 * http://www.apache.org/licenses/LICENSE-2.0 11 * 12 * Unless required by applicable law or agreed to in writing, software 13 * distributed under the License is distributed on an "AS IS" BASIS, 14 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 15 * See the License for the specific language governing permissions and 16 * limitations under the License. 17 */ 18 19 'use strict'; 20 21 // MODULES // 22 23 var float64ToFloat32 = require( '@stdlib/number/float64/base/to-float32' ); 24 var abs = require( '@stdlib/math/base/special/abs' ); 25 26 27 // MAIN // 28 29 /** 30 * Computes the sum of single-precision floating-point strided array elements using a second-order iterative Kahan–Babuška algorithm. 31 * 32 * ## Method 33 * 34 * - This implementation uses a second-order iterative Kahan–Babuška algorithm, as described by Klein (2005). 35 * 36 * ## References 37 * 38 * - Klein, Andreas. 2005. "A Generalized Kahan-Babuška-Summation-Algorithm." _Computing_ 76 (3): 279–93. doi:[10.1007/s00607-005-0139-x](https://doi.org/10.1007/s00607-005-0139-x). 39 * 40 * @param {PositiveInteger} N - number of indexed elements 41 * @param {Float32Array} x - input array 42 * @param {integer} stride - stride length 43 * @returns {number} sum 44 * 45 * @example 46 * var Float32Array = require( '@stdlib/array/float32' ); 47 * 48 * var x = new Float32Array( [ 1.0, -2.0, 2.0 ] ); 49 * var N = x.length; 50 * 51 * var v = ssumkbn2( N, x, 1 ); 52 * // returns 1.0 53 */ 54 function ssumkbn2( N, x, stride ) { 55 var sum; 56 var ccs; 57 var ix; 58 var cs; 59 var cc; 60 var v; 61 var t; 62 var c; 63 var i; 64 65 if ( N <= 0 ) { 66 return 0.0; 67 } 68 if ( N === 1 || stride === 0 ) { 69 return x[ 0 ]; 70 } 71 if ( stride < 0 ) { 72 ix = (1-N) * stride; 73 } else { 74 ix = 0; 75 } 76 sum = 0.0; 77 ccs = 0.0; // second order correction term for lost low order bits 78 cs = 0.0; // first order correction term for lost low order bits 79 for ( i = 0; i < N; i++ ) { 80 v = x[ ix ]; 81 t = float64ToFloat32( sum+v ); 82 if ( abs( sum ) >= abs( v ) ) { 83 c = float64ToFloat32( float64ToFloat32( sum-t ) + v ); 84 } else { 85 c = float64ToFloat32( float64ToFloat32( v-t ) + sum ); 86 } 87 sum = t; 88 t = float64ToFloat32( cs+c ); 89 if ( abs( cs ) >= abs( c ) ) { 90 cc = float64ToFloat32( float64ToFloat32( cs-t ) + c ); 91 } else { 92 cc = float64ToFloat32( float64ToFloat32( c-t ) + cs ); 93 } 94 cs = t; 95 ccs = float64ToFloat32( ccs+cc ); 96 ix += stride; 97 } 98 return float64ToFloat32( sum + float64ToFloat32( cs+ccs ) ); 99 } 100 101 102 // EXPORTS // 103 104 module.exports = ssumkbn2;