102 lines
3.2 KiB
C
102 lines
3.2 KiB
C
/**
|
||
* @license Apache-2.0
|
||
*
|
||
* Copyright (c) 2020 The Stdlib Authors.
|
||
*
|
||
* Licensed under the Apache License, Version 2.0 (the "License");
|
||
* you may not use this file except in compliance with the License.
|
||
* You may obtain a copy of the License at
|
||
*
|
||
* http://www.apache.org/licenses/LICENSE-2.0
|
||
*
|
||
* Unless required by applicable law or agreed to in writing, software
|
||
* distributed under the License is distributed on an "AS IS" BASIS,
|
||
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||
* See the License for the specific language governing permissions and
|
||
* limitations under the License.
|
||
*/
|
||
|
||
#include "stdlib/stats/base/svariancewd.h"
|
||
#include <stdint.h>
|
||
|
||
/**
|
||
* Computes the variance of a single-precision floating-point strided array using Welford's algorithm.
|
||
*
|
||
* ## Method
|
||
*
|
||
* - This implementation uses Welford's algorithm for efficient computation, which can be derived as follows. Let
|
||
*
|
||
* ```tex
|
||
* \begin{align*}
|
||
* S_n &= n \sigma_n^2 \\
|
||
* &= \sum_{i=1}^{n} (x_i - \mu_n)^2 \\
|
||
* &= \biggl(\sum_{i=1}^{n} x_i^2 \biggr) - n\mu_n^2
|
||
* \end{align*}
|
||
* ```
|
||
*
|
||
* Accordingly,
|
||
*
|
||
* ```tex
|
||
* \begin{align*}
|
||
* S_n - S_{n-1} &= \sum_{i=1}^{n} x_i^2 - n\mu_n^2 - \sum_{i=1}^{n-1} x_i^2 + (n-1)\mu_{n-1}^2 \\
|
||
* &= x_n^2 - n\mu_n^2 + (n-1)\mu_{n-1}^2 \\
|
||
* &= x_n^2 - \mu_{n-1}^2 + n(\mu_{n-1}^2 - \mu_n^2) \\
|
||
* &= x_n^2 - \mu_{n-1}^2 + n(\mu_{n-1} - \mu_n)(\mu_{n-1} + \mu_n) \\
|
||
* &= x_n^2 - \mu_{n-1}^2 + (\mu_{n-1} - x_n)(\mu_{n-1} + \mu_n) \\
|
||
* &= x_n^2 - \mu_{n-1}^2 + \mu_{n-1}^2 - x_n\mu_n - x_n\mu_{n-1} + \mu_n\mu_{n-1} \\
|
||
* &= x_n^2 - x_n\mu_n - x_n\mu_{n-1} + \mu_n\mu_{n-1} \\
|
||
* &= (x_n - \mu_{n-1})(x_n - \mu_n) \\
|
||
* &= S_{n-1} + (x_n - \mu_{n-1})(x_n - \mu_n)
|
||
* \end{align*}
|
||
* ```
|
||
*
|
||
* where we use the identity
|
||
*
|
||
* ```tex
|
||
* x_n - \mu_{n-1} = n (\mu_n - \mu_{n-1})
|
||
* ```
|
||
*
|
||
* ## References
|
||
*
|
||
* - Welford, B. P. 1962. "Note on a Method for Calculating Corrected Sums of Squares and Products." _Technometrics_ 4 (3). Taylor & Francis: 419–20. doi:[10.1080/00401706.1962.10490022](https://doi.org/10.1080/00401706.1962.10490022).
|
||
* - van Reeken, A. J. 1968. "Letters to the Editor: Dealing with Neely's Algorithms." _Communications of the ACM_ 11 (3): 149–50. doi:[10.1145/362929.362961](https://doi.org/10.1145/362929.362961).
|
||
*
|
||
* @param N number of indexed elements
|
||
* @param correction degrees of freedom adjustment
|
||
* @param X input array
|
||
* @param stride stride length
|
||
* @return output value
|
||
*/
|
||
float stdlib_strided_svariancewd( const int64_t N, const float correction, const float *X, const int64_t stride ) {
|
||
float delta;
|
||
int64_t ix;
|
||
int64_t i;
|
||
double n;
|
||
float mu;
|
||
float M2;
|
||
float v;
|
||
|
||
n = (double)N - (double)correction;
|
||
if ( N <= 0 || n <= 0.0f ) {
|
||
return 0.0f / 0.0f; // NaN
|
||
}
|
||
if ( N == 1 || stride == 0 ) {
|
||
return 0.0f;
|
||
}
|
||
if ( stride < 0 ) {
|
||
ix = (1-N) * stride;
|
||
} else {
|
||
ix = 0;
|
||
}
|
||
M2 = 0.0f;
|
||
mu = 0.0f;
|
||
for ( i = 0; i < N; i++ ) {
|
||
v = X[ ix ];
|
||
delta = v - mu;
|
||
mu += (float)((double)delta / (double)(i+1));
|
||
M2 += delta * ( v - mu );
|
||
ix += stride;
|
||
}
|
||
return (double)M2 / n;
|
||
}
|