PickLabel#

class dpeeg.transforms.PickLabel(pick: ndarray, keys: list[str] | None = None, order: bool = True, shuffle: bool = True, seed: int = 42)[source]#

Pick a subset of data.

Pick the required labels and data from the dataset and re-label them.

Parameters:

  • pick (ndarray) – Label to include.

  • keys (list of str, optional) – The key of the eegdata value to be transformed, if required. Applies to all eegdata by default.

  • order (bool) – If True, relabel the selected labels.

  • shuffle (bool) – Whether or not to shuffle the data after picking.

  • seed (int) – Controls the shuffling applied to the data after picking.

Returns:

data – Transformed eegdata.

Return type:

eegdata or dataset

Examples

>>> eegdata = dpeeg.EEGData(edata=np.random.randn(16, 3, 10),
...                         label=np.random.randint(0, 3, 16))
array([1, 2, 0, 2, 1, 2, 0, 1, 0, 0, 0, 1, 2, 1, 0, 0])

>>> transforms.PickLabel(np.array([1, 2]))(eegdata, verbose=False)
array([1, 0, 1, 0, 1, 0, 0, 0, 1])

If some values do not need to be transformed, they can be excluded by the keys parameter:

>>> eegdata = dpeeg.EEGData(
...     edata=np.random.randn(16, 3, 10),
...     label=np.random.randint(0, 3, 16),
...     adj=np.random.randn(16, 3, 3),
...     pcc=np.random.randn(16, 3, 3),
... )
>>> transforms.PickLabel(
...    np.array([0, 1]), keys=["edata", "adj"]
... )(eegdata, verbose=False)
[edata=(12, 3, 10), label=(12,), adj=(12, 3, 3), pcc=(16, 3, 3)]