Settings record for the INVLIB::InvCMNE solver (model path, sliding-window length, regularisation, base method). More...


Go to the source code of this file.
Classes | |
| struct | INVLIB::InvCMNESettings |
| CMNE settings. More... | |
Namespaces | |
| namespace | INVLIB |
| Inverse source estimation (MNE, dSPM, sLORETA, dipole fitting). | |
Settings record for the INVLIB::InvCMNE solver (model path, sliding-window length, regularisation, base method).
SPDX-License-Identifier: BSD-3-Clause Copyright (c) 2026 MNE-CPP Authors
INVLIB::InvCMNESettings collects every knob that InvCMNE::compute and InvCMNE::trainLstm need: the path to the trained ONNX LSTM, the look-back window length k that defines the temporal context fed into the network, the source-space size (which must match the LSTM's input dimension), the Tikhonov regulariser \( \lambda^{2} \), the underlying linear method (0 = MNE, 1 = dSPM, 2 = sLORETA) and the loose-orientation constraint.
Definition in file inv_cmne_settings.h.