mescla.types.EMMAResult

class mescla.types.EMMAResult(eigenvalues, loadings, k, scores, fitted, center, scale, species, index, standardize=True)

Bases: object

Outcome of an End-Member Mixing Analysis.

Variables:

• eigenvalues (ndarray, shape (ns,)) – Eigenvalues of the correlation (or covariance) matrix, descending.

• loadings (ndarray, shape (ns, ns)) – Eigenvectors as columns – the weight of each species on each component.

• k (int) – Retained rank: the dimension of the mixing subspace.

• n_endmembers (int) – k + 1 – a k-dimensional simplex needs k+1 vertices (Christophersen & Hooper, 1992).

• scores (ndarray, shape (n_samples, k)) – Sample coordinates in U-space.

• fitted (ndarray, shape (n_samples, ns)) – Concentrations predicted by the rank-k mixing model, in original units.

• scale (center,) – Standardisation applied to the mixed samples. End-members must be projected with these same values.

Parameters:

• eigenvalues (ndarray)

• loadings (ndarray)

• k (int)

• scores (ndarray)

• fitted (ndarray)

• center (ndarray)

• scale (ndarray)

• species (tuple[str, ...])

• index (tuple[str, ...])

• standardize (bool)

eigenvalues: ndarray

loadings: ndarray

k: int

scores: ndarray

fitted: ndarray

center: ndarray

scale: ndarray

species: tuple[str, ...]

index: tuple[str, ...]

standardize: bool = True

property n_endmembers: int

property components: ndarray

The retained eigenvectors, shape (ns, k).

property explained_variance_ratio: ndarray

loadings_frame()

Eigenvector loadings as a DataFrame (species x components).

Return type:

pd.DataFrame

scores_frame()

Return type:

pd.DataFrame