pyriemann.datasets.RandomOverSampler¶

class pyriemann.datasets.RandomOverSampler(metric='riemann', sampling_strategy='auto', random_state=None, n_jobs=1)¶

Random over-sampling for SPD/HPD matrices.

For each class, output SPD/HPD matrices are interpolated along the geodesic between input SPD/HPD matrices [1].

Parameters:

metricstring, default=”riemann”

Metric used for SPD/HPD matrices interpolation (see pyriemann.utils.geodesic.geodesic()).

sampling_strategystr, default=”auto”

Specify the class targeted by the resampling. The number of matrices in the different classes will be equalized. Possible choices are:

“minority”: resample only the minority class;
“not minority”: resample all classes but the minority class;
“not majority”: resample all classes but the majority class;
“all”: resample all classes;
“auto”: equivalent to “not majority”.

random_stateint | RandomState instance | None, default=None

Pass an int for reproducible output across multiple function calls.

n_jobsint, default=1

Number of jobs to use for the computation. This works by computing each of the class resampling in parallel. If -1 all CPUs are used. If 1 is given, no parallel computing code is used at all, which is useful for debugging. For n_jobs below -1, (n_cpus + 1 + n_jobs) are used. Thus for n_jobs = -2, all CPUs but one are used.

Notes

Added in version 0.10.

References

[1]

Data augmentation in Riemannian space for brain-computer interfaces E. Kalunga, S. Chevallier and Q. Barthélemy. ICML Workshop on Statistics, Machine Learning and Neuroscience, 2015.

__init__(metric='riemann', sampling_strategy='auto', random_state=None, n_jobs=1)¶: Init.

fit(X, y)¶

Check parameters of the sampler.

You should use fit_resample in all cases.

Parameters:

Xndarray, shape (n_matrices, n_channels, n_channels): Set of SPD/HPD matrices.
yndarray, shape (n_matrices,): Labels for each matrix.

Returns:

selfobject: Return the instance itself.

fit_resample(X, y)¶

Resample the matrices.

Parameters:

Xndarray, shape (n_matrices, n_channels, n_channels): Set of SPD/HPD matrices.
yndarray, shape (n_matrices,): Labels for each matrix.

Returns:

X_resampledndarray, shape (n_matrices_new, n_channels, n_channels): Set of resampled SPD/HPD matrices.
y_resampledndarray, shape (n_matrices_new,): Labels for each resampled matrix.

get_metadata_routing()¶

Get metadata routing of this object.

Please check User Guide on how the routing mechanism works.

Returns:

routingMetadataRequest: A MetadataRequest encapsulating routing information.

get_params(deep=True)¶

Get parameters for this estimator.

Parameters:

deepbool, default=True: If True, will return the parameters for this estimator and contained subobjects that are estimators.

Returns:

paramsdict: Parameter names mapped to their values.

set_params(**params)¶

Set the parameters of this estimator.

The method works on simple estimators as well as on nested objects (such as Pipeline). The latter have parameters of the form <component>__<parameter> so that it’s possible to update each component of a nested object.

Parameters:

**paramsdict: Estimator parameters.

Returns:

selfestimator instance: Estimator instance.