Press n or j to go to the next uncovered block, b, p or k for the previous block.
| 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 | 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x | /**
* @license Apache-2.0
*
* Copyright (c) 2018 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.
*/
'use strict';
// MODULES //
var closestCentroid = require( './find_closest_centroid.js' );
var updateCentroid = require( './update_centroid.js' );
// MAIN //
/**
* Initializes clusters and associated statistics given a set of centroids.
*
* ## Notes
*
* - We follow the same approach when calculating cluster statistics as if the centroids had been provided by a user (i.e., not computed internally), as this ensures consistency with how statistics are computed when subsequent data vectors are provided to the accumulator.
*
* @private
* @param {ndarray} data - matrix containing data points
* @param {ndarray} centroids - matrix containing centroids
* @param {ndarray} stats - matrix containing cluster statistics
* @param {Function} acc - cluster statistics accumulator
* @param {Function} dist - distance function
*/
function clusters( data, centroids, stats, acc, dist ) {
var ndims;
var cbuf;
var dbuf;
var npts;
var sc;
var sd;
var oc;
var od;
var N;
var k;
var c;
var d;
var i;
k = centroids.shape[ 0 ];
ndims = centroids.shape[ 1 ];
npts = data.shape[ 0 ];
cbuf = centroids.data;
sc = centroids.strides[ 0 ];
dbuf = data.data;
sd = data.strides[ 0 ];
od = 0;
for ( i = 0; i < npts; i++ ) {
// Find the closest centroid by computing the distance from the provided data point to each centroid:
c = closestCentroid( dist, k, ndims, cbuf, sc, 0, dbuf, 1, od ); // Magic numbers arise from knowing that matrices are row-major single-segment contiguous
// Compute the centroids buffer index offset to point to the closest centroid:
oc = sc * c;
// Update the closest centroid:
N = stats.get( c, 0 ) + 1;
updateCentroid( ndims, N, cbuf, 1, oc, dbuf, 1, od ); // Magic number `1` as we know that these matrices are row-major single-segment contiguous
// Recompute the distance based on the updated centroid position:
d = dist( ndims, cbuf, 1, oc, dbuf, 1, od ); // Magic number `1` as we know that these matrices are row-major single-segment contiguous
// Update cluster statistics:
acc( c, d );
// Increment the data buffer index offset to point to the next data point:
od += sd;
}
}
// EXPORTS //
module.exports = clusters;
|