Add match_all method in paradigm to support CompoundDataset evaluation with MNE epochs

docs/source/whats_new.rst

@@ -48,6 +48,7 @@ Enhancements
 - Add :obj:`moabb.datasets.compound_dataset.utils.compound_dataset_list`  (:gh:`455` by `Pierre Guetschel`_)
 - Add c-VEP paradigm and Thielen2021 c-VEP dataset (:gh:`463` by `Jordy Thielen`_)
 - Add option to plot scores vertically. (:gh:`417` by `Sara Sedlar`_)
+- Add match_all method in paradigm to support CompoundDataset evaluation with MNE epochs (:gh:`473` by `Gregoire Cattan`_)
 
 Bugs
 ~~~~

moabb/datasets/fake.py

@@ -45,8 +45,10 @@ def __init__(
         paradigm="imagery",
         channels=("C3", "Cz", "C4"),
         seed=None,
+        sfreq=128,
     ):
         self.n_runs = n_runs
+        self.sfreq = sfreq
         event_id = {ev: ii + 1 for ii, ev in enumerate(event_list)}
         self.channels = channels
         self.seed = seed
@@ -81,7 +83,7 @@ def _get_single_subject_data(self, subject):
 
     def _generate_raw(self):
         montage = make_standard_montage("standard_1005")
-        sfreq = 128
+        sfreq = self.sfreq
         duration = len(self.event_id) * 60
         eeg_data = 2e-5 * np.random.randn(duration * sfreq, len(self.channels))
         y = np.zeros((duration * sfreq))

moabb/paradigms/base.py

@@ -9,6 +9,7 @@
 from sklearn.pipeline import Pipeline, make_pipeline
 from sklearn.preprocessing import FunctionTransformer
 
+from moabb.datasets.base import BaseDataset
 from moabb.datasets.bids_interface import StepType
 from moabb.datasets.preprocessing import (
     EpochsToEvents,
@@ -420,6 +421,43 @@ def _get_array_pipeline(
             return None
         return Pipeline(steps)
 
+    def match_all(self, datasets: List[BaseDataset], shift=-0.5):
+        """
+        Initialize this paradigm to match all datasets in parameter:
+        - `self.resample` is set to match the minimum frequency in all datasets, minus `shift`.
+          If the frequency is 128 for example, then MNE can return 128 or 129 samples
+          depending on the dataset, even if the length of the epochs is 1s
+          Setting `shift=-0.5` solves this particular issue.
+        - `self.channels` is initialized with the channels which are common to all datasets.
+
+        Parameters
+        ----------
+        datasets: List[BaseDataset]
+            A dataset instance.
+        """
+        resample = None
+        channels = None
+        for dataset in datasets:
+            X, _, _ = self.get_data(
+                dataset, subjects=[dataset.subject_list[0]], return_epochs=True
+            )
+            info = X.info
+            sfreq = info["sfreq"]
+            ch_names = info["ch_names"]
+            # get the minimum sampling frequency between all datasets
+            resample = sfreq if resample is None else min(resample, sfreq)
+            # get the channels common to all datasets
+            channels = (
+                set(ch_names)
+                if channels is None
+                else set(channels).intersection(ch_names)
+            )
+        # If resample=128 for example, then MNE can returns 128 or 129 samples
+        # depending on the dataset, even if the length of the epochs is 1s
+        # `shift=-0.5` solves this particular issue.
+        self.resample = resample + shift
+        self.channels = list(channels)
+
     @abc.abstractmethod
     def _get_events_pipeline(self, dataset):
         pass

moabb/tests/paradigms.py

@@ -345,6 +345,35 @@ def used_events(self, dataset):
 
 
 class Test_P300(unittest.TestCase):
+    def test_match_all(self):
+        # Note: the match all property is implemented in the base paradigm.
+        # Thus, although it is located in the P300 section, this test stands for all paradigms.
+        paradigm = SimpleP300()
+        dataset1 = FakeDataset(
+            paradigm="p300",
+            event_list=["Target", "NonTarget"],
+            channels=("C3", "Cz", "Fz"),
+            sfreq=64,
+        )
+        dataset2 = FakeDataset(
+            paradigm="p300",
+            event_list=["Target", "NonTarget"],
+            channels=["C3", "C4", "Cz"],
+            sfreq=256,
+        )
+        dataset3 = FakeDataset(
+            paradigm="p300",
+            event_list=["Target", "NonTarget"],
+            channels=["C3", "Cz", "Fz", "C4"],
+            sfreq=512,
+        )
+        shift = -0.5
+        paradigm.match_all([dataset1, dataset2, dataset3], shift=shift)
+        # match_all should returns the smallest frequency minus 0.5.
+        # See comment inside the match_all method
+        self.assertEqual(paradigm.resample, 64 + shift)
+        self.assertEqual(paradigm.channels.sort(), ["C3", "Cz"].sort())
+
     def test_BaseP300_paradigm(self):
         paradigm = SimpleP300()
         dataset = FakeDataset(paradigm="p300", event_list=["Target", "NonTarget"])