pyriemann.datasets.make_outliers¶
- pyriemann.datasets.make_outliers(n_matrices, mean, sigma, outlier_coeff=10, random_state=None)¶
Generate a set of outlier points.
Simulate data points that are outliers for a given Riemannian Gaussian distribution with fixed mean and dispersion.
- Parameters:
- n_matricesint
How many matrices to generate.
- meanndarray, shape (n_dim, n_dim)
Center of the Riemannian Gaussian distribution.
- sigmafloat
Dispersion of the Riemannian Gaussian distribution.
- outlier_coeff: float, default=10
Coefficient determining how to define an outlier data point, i.e. how many times the sigma parameter its distance to the mean should be.
- random_stateint, RandomState instance or None, default=None
Pass an int for reproducible output across multiple function calls.
- Returns:
- outliersndarray, shape (n_matrices, n_dim, n_dim)
Set of simulated outlier matrix.
Notes
New in version 0.3.