fix: resolve merge conflicts with base autoloop branch

Accept the base branch's version of 43 conflicted files to resolve
merge conflicts and unblock PR merge.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Author: github-actions[bot]

src/bicluster/bicluster.ts

@@ -27,12 +27,13 @@ function svd2(
     for (let _iter = 0; _iter < 30; _iter++) {
       const u = new Float64Array(m);
       for (let i = 0; i < m; i++) {
-        for (let j = 0; j < n; j++) u[i] += (matrix[i]?.[j] ?? 0) * (v[j] ?? 0);
+        for (let j = 0; j < n; j++)
+          u[i]! += (matrix[i]?.[j] ?? 0) * (v[j] ?? 0);
       }
       const newV = new Float64Array(n);
       for (let i = 0; i < m; i++) {
         for (let j = 0; j < n; j++)
-          newV[j] += (matrix[i]?.[j] ?? 0) * (u[i] ?? 0);
+          newV[j]! += (matrix[i]?.[j] ?? 0) * (u[i] ?? 0);
       }
       let norm = 0;
       for (let j = 0; j < n; j++) norm += (newV[j] ?? 0) ** 2;
@@ -42,7 +43,7 @@ function svd2(
     }
     const u = new Float64Array(m);
     for (let i = 0; i < m; i++) {
-      for (let j = 0; j < n; j++) u[i] += (matrix[i]?.[j] ?? 0) * (v[j] ?? 0);
+      for (let j = 0; j < n; j++) u[i]! += (matrix[i]?.[j] ?? 0) * (v[j] ?? 0);
     }
     let sigma = 0;
     for (let i = 0; i < m; i++) sigma += (u[i] ?? 0) ** 2;
@@ -95,12 +96,12 @@ function kmeansSimple(X: Float64Array[], k: number, maxIter = 100): Int32Array {
     for (let j = 0; j < k; j++) centers[j] = new Float64Array(d);
     for (let i = 0; i < n; i++) {
       const c = labels[i]!;
-      counts[c]++;
+      counts[c]!++;
       for (let l = 0; l < d; l++) centers[c]![l]! += X[i]?.[l] ?? 0;
     }
     for (let j = 0; j < k; j++) {
       if ((counts[j] ?? 0) > 0) {
-        for (let l = 0; l < d; l++) centers[j]![l]! /= counts[j];
+        for (let l = 0; l < d; l++) centers[j]![l]! /= counts[j] ?? 1;
       }
     }
   }
@@ -196,8 +197,8 @@ export class SpectralCoclustering {
     const colSums = new Float64Array(nCols);
     for (let i = 0; i < nRows; i++) {
       for (let j = 0; j < nCols; j++) {
-        rowSums[i] += X[i]?.[j] ?? 0;
-        colSums[j] += X[i]?.[j] ?? 0;
+        rowSums[i]! += X[i]?.[j] ?? 0;
+        colSums[j]! += X[i]?.[j] ?? 0;
       }
     }
     const normalized = X.map((row, i) => {

src/cluster/hdbscan.ts

@@ -123,7 +123,7 @@ export class HDBSCAN {
           !inMST[v] &&
           mrd[u]![v]! < (minEdge[v] ?? Number.POSITIVE_INFINITY)
         ) {
-          minEdge[v]! = mrd[u]![v];
+          minEdge[v] = mrd[u]![v] ?? 0;
           parent[v]! = u;
         }
       }
@@ -137,7 +137,7 @@ export class HDBSCAN {
     const find = (x: number): number => {
       let cur = x;
       while (uf[cur] !== cur) {
-        uf[cur]! = uf[uf[cur]!];
+        uf[cur] = uf[uf[cur]!] ?? cur;
         cur = uf[cur]!;
       }
       return cur;
@@ -177,7 +177,7 @@ export class HDBSCAN {
       const sz = clusterSizes[root] ?? 1;
       if (sz >= this.minClusterSize) {
         if (!rootToCluster.has(root)) rootToCluster.set(root, nextCluster++);
-        labels[i]! = rootToCluster.get(root);
+        labels[i] = rootToCluster.get(root) ?? -1;
       }
     }
 

src/covariance/mcd.ts

@@ -103,7 +103,7 @@ export class MinCovDet {
     const n = indices.length;
     const mean = new Float64Array(p);
     for (const idx of indices)
-      for (let j = 0; j < p; j++) mean[j] += (X[idx]![j] ?? 0) / n;
+      for (let j = 0; j < p; j++) mean[j]! += (X[idx]![j] ?? 0) / n;
     const cov: Float64Array[] = Array.from(
       { length: p },
       () => new Float64Array(p),
@@ -113,7 +113,7 @@ export class MinCovDet {
       for (let j = 0; j < p; j++) diff[j] = (X[idx]![j] ?? 0) - (mean[j] ?? 0);
       for (let i = 0; i < p; i++)
         for (let j = 0; j < p; j++)
-          cov[i]![j] += ((diff[i] ?? 0) * (diff[j] ?? 0)) / (n - 1);
+          cov[i]![j]! += ((diff[i] ?? 0) * (diff[j] ?? 0)) / (n - 1);
     }
     return { mean, cov };
   }

src/covariance/shrinkage.ts

@@ -45,10 +45,10 @@ function centerMatrix(
 ): Float64Array {
   const means = new Float64Array(nFeatures);
   for (let i = 0; i < nSamples; i++)
-    for (let j = 0; j < nFeatures; j++) means[j]! += X[i * nFeatures + j];
+    for (let j = 0; j < nFeatures; j++) means[j]! += X[i * nFeatures + j] ?? 0;
   for (let j = 0; j < nFeatures; j++) means[j]! /= nSamples;
   for (let i = 0; i < nSamples; i++)
-    for (let j = 0; j < nFeatures; j++) X[i * nFeatures + j]! -= means[j];
+    for (let j = 0; j < nFeatures; j++) X[i * nFeatures + j]! -= means[j] ?? 0;
   return means;
 }
 

src/cross_decomposition/pls.ts

@@ -204,8 +204,6 @@ export class PLSRegression {
           qy[j] = (qy[j] ?? 0) + (yi[j] ?? 0) * (t[i] ?? 0);
         }
       }
-      if (sNorm2 > 1e-15)
-        for (let j = 0; j < q; j++) qy[j] = (qy[j] ?? 0) / sNorm2;
 
       this.xWeights_[comp] = u;
       this.yWeights_[comp] = v;
@@ -220,13 +218,13 @@ export class PLSRegression {
       Xc = Xc.map((xi, i) => {
         const out = new Float64Array(p);
         for (let j = 0; j < p; j++)
-          out[j] = (xi[j] ?? 0) - (tFull[i] ?? 0) * (px[j] ?? 0);
+          out[j] = (xi[j] ?? 0) - (t[i] ?? 0) * (px[j] ?? 0);
         return out;
       });
       Yc = Yc.map((yi, i) => {
         const out = new Float64Array(q);
         for (let j = 0; j < q; j++)
-          out[j] = (yi[j] ?? 0) - (tFull[i] ?? 0) * (qy[j] ?? 0);
+          out[j] = (yi[j] ?? 0) - (t[i] ?? 0) * (qy[j] ?? 0);
         return out;
       });
     }

src/cross_decomposition/pls_svd.ts

@@ -122,15 +122,15 @@ export class PLSSVDExt {
         const v = new Float64Array(nTargets);
         for (let j = 0; j < nFeatures; j++)
           for (let l = 0; l < nTargets; l++)
-            v[l] += (C[j]?.[l] ?? 0) * (u[j] ?? 0);
+            v[l]! += (C[j]?.[l] ?? 0) * (u[j] ?? 0);
         const normV = Math.sqrt(v.reduce((s, v2) => s + v2 ** 2, 0)) || 1;
         for (let l = 0; l < nTargets; l++) v[l] = (v[l] ?? 0) / normV;
 
         // u = C v
         const uNew = new Float64Array(nFeatures);
         for (let j = 0; j < nFeatures; j++)
           for (let l = 0; l < nTargets; l++)
-            uNew[j] += (C[j]?.[l] ?? 0) * (v[l] ?? 0);
+            uNew[j]! += (C[j]?.[l] ?? 0) * (v[l] ?? 0);
 
         // Orthogonalize against previous
         for (const pu of xWeights) {
@@ -149,7 +149,7 @@ export class PLSSVDExt {
       const v = new Float64Array(nTargets);
       for (let j = 0; j < nFeatures; j++)
         for (let l = 0; l < nTargets; l++)
-          v[l] += (C[j]?.[l] ?? 0) * (u[j] ?? 0);
+          v[l]! += (C[j]?.[l] ?? 0) * (u[j] ?? 0);
       const normV = Math.sqrt(v.reduce((s, v2) => s + v2 ** 2, 0)) || 1;
       for (let l = 0; l < nTargets; l++) v[l] = (v[l] ?? 0) / normV;
 

src/datasets/openml.ts

@@ -64,7 +64,7 @@ export async function fetchOpenML(
   // Parse the dataset list to find the actual dataset ID
   let actualDataId = dataId;
   if (actualDataId == null) {
-    const datasets = json.data as { dataset?: { did?: number }[] } | undefined;
+    const datasets = json.data;
     const did = datasets?.dataset?.[0]?.did;
     if (did == null)
       throw new Error(`fetchOpenML: dataset "${name}" not found`);

src/decomposition/ica.ts

@@ -101,7 +101,7 @@ export class FastICA {
         for (let j = 0; j < p; j++) {
           for (let l = j; l < p; l++) {
             cov[j]![l]! += (Xc[i]![j] ?? 0) * (Xc[i]![l] ?? 0);
-            if (l !== j) cov[l]![j]! = cov[j]![l];
+            if (l !== j) cov[l]![j] = cov[j]![l] ?? 0;
           }
         }
       }

src/decomposition/online_pca.ts

@@ -109,12 +109,12 @@ export class IncrementalPCAOnline {
         const u = new Float64Array(rows);
         for (let i = 0; i < rows; i++) {
           for (let j = 0; j < cols; j++)
-            u[i] += (augmented[i]?.[j] ?? 0) * (V[c]?.[j] ?? 0);
+            u[i]! += (augmented[i]?.[j] ?? 0) * (V[c]?.[j] ?? 0);
         }
         const vNew = new Float64Array(cols);
         for (let i = 0; i < rows; i++) {
           for (let j = 0; j < cols; j++)
-            vNew[j] += (augmented[i]?.[j] ?? 0) * (u[i] ?? 0);
+            vNew[j]! += (augmented[i]?.[j] ?? 0) * (u[i] ?? 0);
         }
         // Orthogonalize against previous
         for (let p = 0; p < c; p++) {

src/discriminant_analysis/qda.ts

@@ -59,7 +59,7 @@ export class QuadraticDiscriminantAnalysis {
       // Class mean
       const mean = new Float64Array(nFeatures);
       for (const x of classX) {
-        for (let j = 0; j < nFeatures; j++) mean[j] += (x[j] ?? 0) / nc;
+        for (let j = 0; j < nFeatures; j++) mean[j]! += (x[j] ?? 0) / nc;
       }
       this.means_.push(mean);
 
@@ -71,14 +71,14 @@ export class QuadraticDiscriminantAnalysis {
       for (const x of classX) {
         for (let j = 0; j < nFeatures; j++) {
           for (let k = 0; k < nFeatures; k++) {
-            cov[j]![k] +=
+            cov[j]![k]! +=
               (((x[j] ?? 0) - mean[j]!) * ((x[k] ?? 0) - mean[k]!)) / (nc - 1);
           }
         }
       }
       // Regularization
       if (this.regParam > 0) {
-        for (let j = 0; j < nFeatures; j++) cov[j]![j] += this.regParam;
+        for (let j = 0; j < nFeatures; j++) cov[j]![j]! += this.regParam;
       }
 
       if (this.storeCovariance) this.covariance_.push(cov);