JadeICA¶
-
class
qaa.decomposition.jade.JadeICA(*, n_components=None)¶ Bases:
sklearn.base.TransformerMixin,sklearn.base.BaseEstimatorPerform blind source separation using joint diagonalization.
Methods Summary
fit(arr[, y])Calculate the unmixing matrix.
fit_transform(arr[, y])Determine the independent signals using joint diagonalization.
inverse_transform(arr)Find the original mixed signals.
transform(arr)Project the unmixing matrix onto arr to give independent signals.
Methods Documentation
-
fit(arr, y=None)¶ Calculate the unmixing matrix.
- Parameters
arr (NDArray) – mixed signal array
y (NDArray, optional) – unused
- Returns
- Return type
self
- Raises
IndexError – if \(m > n_features\)
-
fit_transform(arr, y=None, **fit_params)¶ Determine the independent signals using joint diagonalization.
- Parameters
arr (NDArray) – Mixed signal array
y (NDArray, optional) – unused
fit_params (dict) – unused
- Returns
Unmixed signal array
- Return type
NDArray
- Raises
IndexError – if \(m > n_features\)
-
inverse_transform(arr)¶ Find the original mixed signals.
- Parameters
arr (NDArray) – Unmixed signal array
- Raises
NotImplementedError – If called because Jade does not provide inversion
- Return type
nptyping.types._ndarray.NDArray[(typing.Any, Ellipsis), nptyping.types._number.Float[float, numpy.floating]]
-
transform(arr)¶ Project the unmixing matrix onto arr to give independent signals.
- Parameters
arr (NDArray) – Unmixed signals
- Returns
Unmixed signal array
- Return type
NDArray
-