Limitations

SheShe provides tools for modal analysis, but several limitations apply:

Dependence on surface representation

The algorithms assume a smooth and well-sampled surface. Imperfect meshes or sparse points degrade boundary estimates.

Mitigation: preprocess meshes to remove noise, interpolate missing regions and validate against known benchmarks.

High-dimensional approximations

In very high dimensions SheShe relies on approximations that may miss subtle structures and increase computation time.

Mitigation: apply dimensionality reduction or focus on informative subspaces with SubspaceScout.

Noise and sampling sensitivity

Noisy or sparsely sampled data can produce unstable boundaries.

Mitigation: denoise inputs and aggregate multiple runs through ModalScoutEnsemble.

Parameter tuning

Performance depends on choosing suitable bandwidths and clustering thresholds.

Mitigation: explore settings via the Quick API and cross-validation.