Skip to main content

torch.js has not been released yet.

Playground Contact

Introduction Type Safety Tensor Expressions Tensor Indexing Einsum Einops Autograd Training a Model Profiling & Memory PyTorch Migration Best Practices Runtimes Performance PyTorch Compatibility Benchmarks DType Coverage

Introduction Renderer Gallery Renderers Analysis

computeQuantiles computePercentiles computeCorrelation computeCovariance detectOutliers detectOutliersZScore computeExtendedStats describeExtendedStats QuantileResult CorrelationResult OutlierResult ExtendedStats createDBSCAN dbscan DBSCAN DBSCAN.fit DBSCAN.destroy DBSCANOptions DBSCANResult createHierarchicalClustering hierarchicalClustering HierarchicalClustering HierarchicalClustering.fit HierarchicalClustering.destroy LinkageMethod DistanceMetric HierarchicalOptions DendrogramNode HierarchicalResult computeHistogram getHistogramCenters normalizeHistogram HistogramResult createKDE kde KDE KDE.fit KDE.destroy KernelType KDEOptions KDEResult createKMeans kmeans KMeans KMeans.fit KMeans.destroy KMeansOptions KMeansResult createPCA pca PCA PCA.fit PCA.destroy PCAOptions PCAResult throttleProgress consoleProgress ProgressInfo ProgressCallback computeStats computeMinMax describeStats TensorStats createTSNE tsne TSNE TSNE.fit TSNE.destroy TSNEOptions TSNEResult createUMAP umap UMAP UMAP.fit UMAP.destroy UMAPOptions UMAPResult

torch.js· 2026

Legal Terms of Use Privacy Policy

/

/

docs
viz
viz
analysis
detectOutliersZScore

viz.analysis.detectOutliersZScore

function detectOutliersZScore(tensor: Tensor, threshold = 3.0): Promise<{ outlierIndices: number[]; outlierValues: number[]; count: number }>

Detect outliers using Z-score method.

Outliers are values with |z-score| > threshold (default 3.0).

Parameters

tensorTensor: Input tensor
thresholdunknownoptional: Z-score threshold (default 3.0)

Returns

Promise<{ outlierIndices: number[]; outlierValues: number[]; count: number }>– Array of outlier indices