[MNT] Solving moabb and braindecode compatibility

docs/source/whats_new.rst

@@ -24,7 +24,9 @@ Bugs
 - Fix Stieger2021 dataset bugs (:gh:`651` by `Martin Wimpff`_)
 - Unpinning major version Scikit-learn and numpy (:gh:`652` by `Bruno Aristimunha`_)
 - Replacing the func:`numpy.string_` to func:`numpy.bytes_` (:gh:`665` by `Bruno Aristimunha`_)
--  Fixing the set_download_dir that was not working when we tried to set the dir more than 10 times at the same time (:gh:`668` by `Bruno Aristimunha`_)
+- Fixing the set_download_dir that was not working when we tried to set the dir more than 10 times at the same time (:gh:`668` by `Bruno Aristimunha`_)
+- Creating stimulus channels in :class:`moabb.datasets.Zhou2016` and :class:`moabb.datasets.PhysionetMI` to allow braindecode compatibility (:gh:`669` by `Bruno Aristimunha`_)
+
 
 API changes
 ~~~~~~~~~~~

moabb/datasets/Zhou2016.py

@@ -14,6 +14,7 @@
 
 from .base import BaseDataset
 from .download import get_dataset_path
+from .utils import stim_channels_with_selected_ids
 
 
 DATA_PATH = "https://ndownloader.figshare.com/files/3662952"
@@ -88,6 +89,7 @@ def __init__(self):
             paradigm="imagery",
             doi="10.1371/journal.pone.0162657",
         )
+        self.events = dict(left_hand=1, right_hand=2, feet=3)
 
     def _get_single_subject_data(self, subject):
         """Return data for a single subject."""
@@ -105,7 +107,9 @@ def _get_single_subject_data(self, subject):
                 stim[stim == "2"] = "right_hand"
                 stim[stim == "3"] = "feet"
                 raw.annotations.description = stim
-                out[sess_key][run_key] = raw
+                out[sess_key][run_key] = stim_channels_with_selected_ids(
+                    raw, desired_event_id=self.events
+                )
                 out[sess_key][run_key].set_montage(make_standard_montage("standard_1005"))
         return out
 

moabb/datasets/physionet_mi.py

@@ -6,6 +6,7 @@
 
 from moabb.datasets.base import BaseDataset
 from moabb.datasets.download import data_dl, get_dataset_path
+from moabb.datasets.utils import stim_channels_with_selected_ids
 
 
 BASE_URL = "https://physionet.org/files/eegmmidb/1.0.0/"
@@ -79,7 +80,7 @@ def __init__(self, imagined=True, executed=False):
             paradigm="imagery",
             doi="10.1109/TBME.2004.827072",
         )
-
+        self.events = dict(left_hand=2, right_hand=3, feet=5, hands=4, rest=1)
         self.imagined = imagined
         self.executed = executed
         self.feet_runs = []
@@ -123,7 +124,9 @@ def _get_single_subject_data(self, subject):
             stim[stim == "T1"] = "left_hand"
             stim[stim == "T2"] = "right_hand"
             raw.annotations.description = stim
-            data[str(idx)] = raw
+            data[str(idx)] = stim_channels_with_selected_ids(
+                raw, desired_event_id=self.events
+            )
             idx += 1
 
         # feet runs
@@ -136,7 +139,9 @@ def _get_single_subject_data(self, subject):
             stim[stim == "T1"] = "hands"
             stim[stim == "T2"] = "feet"
             raw.annotations.description = stim
-            data[str(idx)] = raw
+            data[str(idx)] = stim_channels_with_selected_ids(
+                raw, desired_event_id=self.events
+            )
             idx += 1
 
         return {"0": data}
@@ -171,3 +176,41 @@ def _load_data(self, subject, runs, path=None, force_update=False, verbose=None)
             p = data_dl(url, sign, path, force_update, verbose)
             data_paths.append(p)
         return data_paths
+
+    def _create_stim_channels(self, raw):
+        # Define a consistent mapping from event descriptions to integer IDs
+        desired_event_id = self.events
+
+        # Get events using the consistent event_id mapping
+        events, _ = mne.events_from_annotations(raw, event_id=desired_event_id)
+
+        # Filter the events array to include only desired events
+        desired_event_ids = list(desired_event_id.values())
+        filtered_events = events[np.isin(events[:, 2], desired_event_ids)]
+
+        # Create annotations from filtered events using the inverted mapping
+        event_desc = {v: k for k, v in desired_event_id.items()}
+        annot_from_events = mne.annotations_from_events(
+            events=filtered_events,
+            event_desc=event_desc,
+            sfreq=raw.info["sfreq"],
+            orig_time=raw.info["meas_date"],
+        )
+        raw.set_annotations(annot_from_events)
+
+        # Create the stim channel data array
+        stim_channs = np.zeros((1, raw.n_times))
+        for event in filtered_events:
+            sample_index = event[0]
+            event_code = event[2]  # Consistent event IDs
+            stim_channs[0, sample_index] = event_code
+
+        # Create the stim channel and add it to raw
+        stim_channel_name = "STIM"
+        stim_info = mne.create_info(
+            [stim_channel_name], sfreq=raw.info["sfreq"], ch_types=["stim"]
+        )
+        stim_raw = mne.io.RawArray(stim_channs, stim_info, verbose=False)
+        raw_with_stim = raw.copy().add_channels([stim_raw], force_update_info=True)
+
+        return raw_with_stim

moabb/datasets/utils.py

@@ -2,6 +2,7 @@
 
 import inspect
 
+import mne
 import numpy as np
 from mne import create_info
 from mne.io import RawArray
@@ -273,3 +274,52 @@ def add_stim_channel_epoch(
     )
     raw = raw.add_channels([RawArray(data=stim_chan, info=info, verbose=False)])
     return raw
+
+
+def stim_channels_with_selected_ids(
+    raw: mne.io.BaseRaw, desired_event_id: dict, stim_channel_name="STIM"
+):
+    """
+    Add a stimulus channel with filtering and renaming based on events_ids.
+
+    Parameters
+    ----------
+    raw: mne.Raw
+        The raw object to add the stimulus channel to.
+    desired_event_id: dict
+        Dictionary with events
+    """
+
+    # Get events using the consistent event_id mapping
+    events, _ = mne.events_from_annotations(raw, event_id=desired_event_id)
+
+    # Filter the events array to include only desired events
+    desired_event_ids = list(desired_event_id.values())
+    filtered_events = events[np.isin(events[:, 2], desired_event_ids)]
+
+    # Create annotations from filtered events using the inverted mapping
+    event_desc = {v: k for k, v in desired_event_id.items()}
+    annot_from_events = mne.annotations_from_events(
+        events=filtered_events,
+        event_desc=event_desc,
+        sfreq=raw.info["sfreq"],
+        orig_time=raw.info["meas_date"],
+    )
+    raw.set_annotations(annot_from_events)
+
+    # Create the stim channel data array
+    stim_channs = np.zeros((1, raw.n_times))
+    for event in filtered_events:
+        sample_index = event[0]
+        event_code = event[2]  # Consistent event IDs
+        stim_channs[0, sample_index] = event_code
+
+    # Create the stim channel and add it to raw
+
+    stim_info = mne.create_info(
+        [stim_channel_name], sfreq=raw.info["sfreq"], ch_types=["stim"]
+    )
+    stim_raw = mne.io.RawArray(stim_channs, stim_info, verbose=False)
+    raw_with_stim = raw.copy().add_channels([stim_raw], force_update_info=True)
+
+    return raw_with_stim