Standardize data across channels. Parameters ---------- epochs_data : array, shape (n_epochs, n_channels[, n_times]) The data. Returns ------- X : array, shape (n_epochs, n_channels, n_times) The data concatenated over channel
(self, epochs_data)
| 209 | return self |
| 210 | |
| 211 | def transform(self, epochs_data): |
| 212 | """Standardize data across channels. |
| 213 | |
| 214 | Parameters |
| 215 | ---------- |
| 216 | epochs_data : array, shape (n_epochs, n_channels[, n_times]) |
| 217 | The data. |
| 218 | |
| 219 | Returns |
| 220 | ------- |
| 221 | X : array, shape (n_epochs, n_channels, n_times) |
| 222 | The data concatenated over channels. |
| 223 | |
| 224 | Notes |
| 225 | ----- |
| 226 | This function makes a copy of the data before the operations and the |
| 227 | memory usage may be large with big data. |
| 228 | """ |
| 229 | check_is_fitted(self, "scaler_") |
| 230 | epochs_data = self._check_data(epochs_data, atleast_3d=False) |
| 231 | if epochs_data.ndim == 2: # can happen with SlidingEstimator |
| 232 | if self.info is not None: |
| 233 | assert len(self.info["ch_names"]) == epochs_data.shape[1] |
| 234 | epochs_data = epochs_data[..., np.newaxis] |
| 235 | assert epochs_data.ndim == 3, epochs_data.shape |
| 236 | return _sklearn_reshape_apply(self.scaler_.transform, True, epochs_data) |
| 237 | |
| 238 | def fit_transform(self, epochs_data, y=None): |
| 239 | """Fit to data, then transform it. |
no test coverage detected