Luxar Viewer API Documentation - v2026.9.22
    Preparing search index...

    Function computeMarginalCholesky

    • Compute the correct marginal Cholesky factor for a subset of dimensions.

      For Σ = L·Lᵀ, the marginal covariance for dimensions S is: Σ_S[i,j] = Σ_k L[s_i,k]·L[s_j,k]

      This function reconstructs Σ_S and then Cholesky-factorizes it. Simply extracting the raw L row/column sub-matrix is INCORRECT when there are cross-dimension correlations.

      Parameters

      • fullPackedL: Float32Array

        Full packed Cholesky factor array

      • fullPackedOffset: number

        Offset into fullPackedL for this splat

      • keepDims: number[] | Uint32Array<ArrayBufferLike>

        Ordered dimension indices to keep; their order defines the output-axis order

      • subNdim: number

        Number of dimensions to keep

      • output: Float32Array

        Output packed marginal Cholesky [subNdim*(subNdim+1)/2]

      • outputOffset: number

        Start offset in output array

      Returns void

    Settings
    Member Visibility