pyriemann.utils.viz.plot_embedding

pyriemann.utils.viz.plot_embedding(X, y=None, *, embd_type='Spectral', metric='riemann', title='Embedding of SPD matrices', normalize=True, max_iter=50)

Plot embedding of SPD matrices.

Parameters:
Xndarray, shape (n_matrices, n_channels, n_channels)

Set of SPD matrices.

yNone | ndarray, shape (n_matrices,), default=None

Labels for each matrix.

embd_type{“Spectral”, “LocallyLinear”, “TSNE”}, default=”Spectral”

Type of the embedding.

metricstring, default=”riemann”

Metric used for the embedding. Can be:

  • “riemann”, “logeuclid”, “euclid”, “logdet”, “kullback”, “kullback_right”, “kullback_sym” for Spectral Embedding;

  • “riemann”, “logeuclid”, “euclid” for Locally Linear Embedding;

  • “riemann”, “logeuclid”, “euclid” for TSNE.

titlestr, default=”Embedding of SPD matrices”

Title of figure.

normalizebool, default=True

If True, the plot is normalized from -1 to +1.

max_iterint, default=50

Maximum number of iterations used for the gradient descent of TSNE.

Returns:
figmatplotlib figure

Figure of embedding.

Notes

Added in version 0.2.6.