time-to-botec

Benchmark sampling in different programming languages
Log | Files | Refs | README

dmeanvarpn.js (3369B)

      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 isnan = require( '@stdlib/math/base/assert/is-nan' );
     24 var dsumpw = require( '@stdlib/blas/ext/base/dsumpw' );
     25 
     26 
     27 // MAIN //
     28 
     29 /**
     30 * Computes the mean and variance of a double-precision floating-point strided array using a two-pass algorithm.
     31 *
     32 * ## Method
     33 *
     34 * -   This implementation uses a two-pass approach, as suggested by Neely (1966).
     35 *
     36 * ## References
     37 *
     38 * -   Neely, Peter M. 1966. "Comparison of Several Algorithms for Computation of Means, Standard Deviations and Correlation Coefficients." _Communications of the ACM_ 9 (7). Association for Computing Machinery: 496–99. doi:[10.1145/365719.365958](https://doi.org/10.1145/365719.365958).
     39 * -   Schubert, Erich, and Michael Gertz. 2018. "Numerically Stable Parallel Computation of (Co-)Variance." In _Proceedings of the 30th International Conference on Scientific and Statistical Database Management_. New York, NY, USA: Association for Computing Machinery. doi:[10.1145/3221269.3223036](https://doi.org/10.1145/3221269.3223036).
     40 *
     41 * @param {PositiveInteger} N - number of indexed elements
     42 * @param {number} correction - degrees of freedom adjustment
     43 * @param {Float64Array} x - input array
     44 * @param {integer} strideX - `x` stride length
     45 * @param {Float64Array} out - output array
     46 * @param {integer} strideOut - `out` stride length
     47 * @returns {Float64Array} output array
     48 *
     49 * @example
     50 * var Float64Array = require( '@stdlib/array/float64' );
     51 *
     52 * var x = new Float64Array( [ 1.0, -2.0, 2.0 ] );
     53 * var out = new Float64Array( 2 );
     54 *
     55 * var v = dmeanvarpn( x.length, 1, x, 1, out, 1 );
     56 * // returns <Float64Array>[ ~0.3333, ~4.3333 ]
     57 */
     58 function dmeanvarpn( N, correction, x, strideX, out, strideOut ) {
     59 	var mu;
     60 	var ix;
     61 	var io;
     62 	var M2;
     63 	var M;
     64 	var d;
     65 	var c;
     66 	var n;
     67 	var i;
     68 
     69 	if ( strideX < 0 ) {
     70 		ix = (1-N) * strideX;
     71 	} else {
     72 		ix = 0;
     73 	}
     74 	if ( strideOut < 0 ) {
     75 		io = -strideOut;
     76 	} else {
     77 		io = 0;
     78 	}
     79 	if ( N <= 0 ) {
     80 		out[ io ] = NaN;
     81 		out[ io+strideOut ] = NaN;
     82 		return out;
     83 	}
     84 	n = N - correction;
     85 	if ( N === 1 || strideX === 0 ) {
     86 		out[ io ] = x[ ix ];
     87 		if ( n <= 0.0 ) {
     88 			out[ io+strideOut ] = NaN;
     89 		} else {
     90 			out[ io+strideOut ] = 0.0;
     91 		}
     92 		return out;
     93 	}
     94 	// Compute an estimate for the mean:
     95 	mu = dsumpw( N, x, strideX ) / N;
     96 	if ( isnan( mu ) ) {
     97 		out[ io ] = NaN;
     98 		out[ io+strideOut ] = NaN;
     99 		return out;
    100 	}
    101 	// Compute the sum of squared differences from the mean...
    102 	M2 = 0.0;
    103 	M = 0.0;
    104 	for ( i = 0; i < N; i++ ) {
    105 		d = x[ ix ] - mu;
    106 		M2 += d * d;
    107 		M += d;
    108 		ix += strideX;
    109 	}
    110 	// Compute an error term for the mean:
    111 	c = M / N;
    112 
    113 	out[ io ] = mu + c;
    114 	if ( n <= 0.0 ) {
    115 		out[ io+strideOut ] = NaN;
    116 	} else {
    117 		out[ io+strideOut ] = (M2/n) - (c*(M/n));
    118 	}
    119 	return out;
    120 }
    121 
    122 
    123 // EXPORTS //
    124 
    125 module.exports = dmeanvarpn;