fix(k-means): use dist metric in initKmeanspp()

- update to use user-provided distance metric

packages/k-means/src/kmeans.ts

@@ -1,15 +1,7 @@
 import { assert } from "@thi.ng/api";
 import { argmin, DIST_SQ, IDistance } from "@thi.ng/distance";
 import { SYSTEM, uniqueIndices, weightedRandom } from "@thi.ng/random";
-import {
-    add,
-    distSq,
-    median,
-    mulN,
-    ReadonlyVec,
-    Vec,
-    zeroes,
-} from "@thi.ng/vectors";
+import { add, median, mulN, ReadonlyVec, Vec, zeroes } from "@thi.ng/vectors";
 import type { CentroidStrategy, Cluster, KMeansOpts } from "./api";
 
 export const kmeans = <T extends ReadonlyVec>(
@@ -25,7 +17,7 @@ export const kmeans = <T extends ReadonlyVec>(
     };
     const num = samples.length;
     const dim = samples[0].length;
-    const centroidIDs = initial || initKmeanspp(k, samples, rnd);
+    const centroidIDs = initial || initKmeanspp(k, samples, dist, rnd);
     assert(centroidIDs.length === k, `wrong number of initial centroids`);
     const centroids: Vec[] = centroidIDs.map((i) => samples[i]);
     const clusters: number[] = [];
@@ -63,20 +55,23 @@ export const kmeans = <T extends ReadonlyVec>(
  *
  * @param k
  * @param samples
+ * @param dist
  * @param rnd
  */
 export const initKmeanspp = <T extends ReadonlyVec>(
     k: number,
     samples: T[],
+    dist: IDistance<ReadonlyVec> = DIST_SQ,
     rnd = SYSTEM
 ) => {
     const num = samples.length;
     const centroidIDs = [rnd.int() % num];
     const centroids = [samples[centroidIDs[0]]];
     const indices = new Array(num).fill(0).map((_, i) => i);
     while (centroidIDs.length < k) {
-        let probs = samples.map((p) =>
-            distSq(p, centroids[argmin(p, centroids)!])
+        let probs = samples.map(
+            (p) =>
+                dist.from(dist.metric(p, centroids[argmin(p, centroids)!])) ** 2
         );
         const id = weightedRandom(indices, probs)();
         centroidIDs.push(id);