Skip to content

Numeric Summary (Eigendecomposition)

This option offers an eigendecomposition analysis for the Eigendecomposition Denoising and Eigendecomposition Filtering options.

Initial Information

The following information is always included:

  • Report date
  • Main, x, and y data titles
  • File source for data
  • Data size
  • Algorithm use for eigendecomposition
  • Model order (embedding dimension)
  • List of eigenmodes reconstructed
  • r² correlation coefficient for reconstruction

Eigenvalues

This section lists the component eigenvalues, the eigenvalues normalized so that each sum to 100%, and the r² correlation coefficient for that component's reconstruction. Both the normalized eigenvalue %, ranging from 100 to 0, and the partial r² value, ranging from 1.0 to 0.0, are indicators of the measure of variance in the original data captured by that eigenmode.

Each eigenvalue and eigenvalue percent is reported, since these are part of the decomposition. Those not used in the reconstruction will be grayed. The partial r² values are reported only for those eigenmodes used in the reconstruction.