evomap.mapping._mds#

Stress-Based Multidimensional Scaling.

Module Contents#

Classes#

MDS

Functions#

_normalized_stress_function(positions, disparities[, ...])

Compute normalized stress as a measure of goodness-of-fit between

_normalized_stress_gradient(positions, distances, ...)

Calculate gradient of normalized stress function.

Attributes#

EPSILON

evomap.mapping._mds.EPSILON = 1e-10#
class evomap.mapping._mds.MDS(n_dims=2, mds_type=None, n_iter=2000, n_iter_check=50, init=None, verbose=0, input_type='distance', max_halves=5, tol=0.001, n_inits=1, step_size=1)[source]#
fit(X)[source]#
fit_transform(X)[source]#
evomap.mapping._mds._normalized_stress_function(positions, disparities, mds_type=None, compute_error=True, compute_grad=True)[source]#

Compute normalized stress as a measure of goodness-of-fit between input distances and the distances among the estimated positions.

Parameters

  • positions (np.array of shape (n_samples, n_dims)) – estimated positions

  • disparities (np.array of shape (n_samples, n_samples)) – input distances (or transform disparities)

  • inclusions (np.array of shape (n_samples), optional) – array of 0/1 entries indicating if an object should be included in the estimation, by default None

  • compute_error (bool, optional) – indicates if cost funciton value should be computed, by default True

  • compute_grad (bool, optional) – indicates if gradient should be computed, by default True

Returns

cost function value and gradient

Return type

float, array of shape (n_samples, n_dims)

evomap.mapping._mds._normalized_stress_gradient(positions, distances, disparities)[source]#

Calculate gradient of normalized stress function.

Parameters

  • positions (np.array of shape (n_samples, n_dims)) – estimated postiions

  • distances (np.array of shape (n_samples, n_samples)) – euclidean distances among estimated positions

  • disparities (np.array of shape (n_samples, n_samples)) – input distance (or disparity) matrix

Returns

gradient

Return type

np.array of shape (n_samples, n_dims)