Braindecode 1.3.0.dev175563314__tar.gz → 1.3.0.dev175887834__tar.gz

{braindecode-1.3.0.dev175563314/braindecode.egg-info → braindecode-1.3.0.dev175887834}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: braindecode
-Version: 1.3.0.dev175563314
+Version: 1.3.0.dev175887834
 Summary: Deep learning software to decode EEG, ECG or MEG signals
 Author-email: Robin Tibor Schirrmeister <robintibor@gmail.com>, Bruno Aristimunha Pinto <b.aristimunha@gmail.com>, Alexandre Gramfort <agramfort@meta.com>
 Maintainer-email: Alexandre Gramfort <agramfort@meta.com>, Bruno Aristimunha Pinto <b.aristimunha@gmail.com>, Robin Tibor Schirrmeister <robintibor@gmail.com>
@@ -22,12 +22,17 @@ Requires-Python: >=3.11
 Description-Content-Type: text/x-rst
 License-File: LICENSE.txt
 License-File: NOTICE.txt
-Requires-Dist: torch>=2.2
-Requires-Dist: torchaudio>=2.0
 Requires-Dist: mne>=1.11.0
 Requires-Dist: mne_bids>=0.16
+Requires-Dist: numpy
+Requires-Dist: pandas
+Requires-Dist: scipy
+Requires-Dist: matplotlib
 Requires-Dist: h5py
 Requires-Dist: skorch>=1.3.0
+Requires-Dist: torch>=2.0
+Requires-Dist: torchaudio>=2.0
+Requires-Dist: einops
 Requires-Dist: joblib
 Requires-Dist: torchinfo
 Requires-Dist: wfdb
@@ -35,7 +40,7 @@ Requires-Dist: linear_attention_transformer
 Requires-Dist: docstring_inheritance
 Requires-Dist: rotary_embedding_torch
 Provides-Extra: moabb
-Requires-Dist: moabb>=1.4.3; extra == "moabb"
+Requires-Dist: moabb>=1.4.2; extra == "moabb"
 Provides-Extra: eegprep
 Requires-Dist: eegprep[eeglabio]>=0.2.23; extra == "eegprep"
 Provides-Extra: hub
@@ -47,7 +52,6 @@ Requires-Dist: pytest-cov; extra == "tests"
 Requires-Dist: codecov; extra == "tests"
 Requires-Dist: pytest_cases; extra == "tests"
 Requires-Dist: mypy; extra == "tests"
-Requires-Dist: transformers>=4.57.0; extra == "tests"
 Provides-Extra: typing
 Requires-Dist: exca==0.4; extra == "typing"
 Requires-Dist: numpydantic>=1.7; extra == "typing"
@@ -71,7 +75,6 @@ Requires-Dist: openneuro-py; extra == "docs"
 Requires-Dist: plotly; extra == "docs"
 Requires-Dist: shap; extra == "docs"
 Requires-Dist: nbformat; extra == "docs"
-Requires-Dist: transformers; extra == "docs"
 Provides-Extra: all
 Requires-Dist: braindecode[moabb]; extra == "all"
 Requires-Dist: braindecode[tests]; extra == "all"

{braindecode-1.3.0.dev175563314 → braindecode-1.3.0.dev175887834}/braindecode/datasets/__init__.py

@@ -11,15 +11,6 @@ from .base import (
 )
 from .bcicomp import BCICompetitionIVDataset4
 from .bids import BIDSDataset, BIDSEpochsDataset
-from .bids_format import (
-    BIDSDerivativesLayout,
-    create_channels_tsv,
-    create_eeg_json_sidecar,
-    create_events_tsv,
-    create_participants_tsv,
-    description_to_bids_path,
-    make_dataset_description,
-)
 from .chb_mit import CHBMIT
 from .mne import create_from_mne_epochs, create_from_mne_raw
 from .moabb import BNCI2014_001, HGD, MOABBDataset
@@ -38,13 +29,6 @@ __all__ = [
     "BaseConcatDataset",
     "BIDSDataset",
     "BIDSEpochsDataset",
-    "BIDSDerivativesLayout",
-    "create_events_tsv",
-    "create_participants_tsv",
-    "create_channels_tsv",
-    "create_eeg_json_sidecar",
-    "make_dataset_description",
-    "description_to_bids_path",
     "MOABBDataset",
     "HGD",
     "BNCI2014_001",

{braindecode-1.3.0.dev175563314 → braindecode-1.3.0.dev175887834}/braindecode/datasets/hub.py

@@ -3,21 +3,9 @@ Hugging Face Hub integration for EEG datasets.
 
 This module provides push_to_hub() and pull_from_hub() functionality
 for braindecode datasets, similar to the model Hub integration.
-
-The format follows a BIDS-like derivatives structure:
-- derivatives/braindecode/
-  - dataset_description.json
-  - participants.tsv
-  - dataset.zarr/ (main data for efficient training)
-  - sub-<label>/
-    - eeg/
-      - *_events.tsv
-      - *_channels.tsv
-      - *_eeg.json
 """
 
 # Authors: Kuntal Kokate
-#          Bruno Aristimunha <b.aristimunha@gmail.com>
 #
 # License: BSD (3-clause)
 
@@ -42,8 +30,7 @@ if TYPE_CHECKING:
 import braindecode
 
 # Import shared validation utilities
-# Import BIDS format utilities
-from . import bids_format, hub_validation
+from . import hub_validation
 
 # Import registry for dynamic class lookup (avoids circular imports)
 from .registry import get_dataset_class, get_dataset_type
@@ -89,15 +76,13 @@ class HubDatasetMixin:
         create_pr: bool = False,
         compression: str = "blosc",
         compression_level: int = 5,
-        pipeline_name: str = "braindecode",
     ) -> str:
         """
-        Upload the dataset to the Hugging Face Hub in BIDS-like Zarr format.
+        Upload the dataset to the Hugging Face Hub in Zarr format.
 
         The dataset is converted to Zarr format with blosc compression, which provides
-        optimal random access performance for PyTorch training. The data is stored
-        in a BIDS derivatives-like structure with events.tsv, channels.tsv,
-        and participants.tsv sidecar files.
+        optimal random access performance for PyTorch training (based on comprehensive
+        benchmarking).
 
         Parameters
         ----------
@@ -115,8 +100,6 @@ class HubDatasetMixin:
             Compression algorithm for Zarr. Options: "blosc", "zstd", "gzip", None.
         compression_level : int, default=5
             Compression level (0-9). Level 5 provides optimal balance.
-        pipeline_name : str, default="braindecode"
-            Name of the processing pipeline for BIDS derivatives.
 
         Returns
         -------
@@ -133,11 +116,18 @@ class HubDatasetMixin:
         Examples
         --------
         >>> dataset = NMT(path=path, preload=True)
-        >>> # Upload with BIDS-like structure
+        >>> # Upload with default settings (zarr with blosc compression)
         >>> url = dataset.push_to_hub(
         ...     repo_id="myusername/nmt-dataset",
         ...     commit_message="Upload NMT EEG dataset"
         ... )
+        >>>
+        >>> # Or customize compression
+        >>> url = dataset.push_to_hub(
+        ...     repo_id="myusername/nmt-dataset",
+        ...     compression="blosc",
+        ...     compression_level=5
+        ... )
         """
         if huggingface_hub is False or zarr is False:
             raise ImportError(
@@ -164,45 +154,25 @@ class HubDatasetMixin:
         with tempfile.TemporaryDirectory() as tmpdir:
             tmp_path = Path(tmpdir)
 
-            # Create BIDS-like derivatives structure
-            log.info("Creating BIDS-like derivatives structure...")
-            bids_layout = bids_format.BIDSDerivativesLayout(
-                tmp_path, pipeline_name=pipeline_name
-            )
-            derivatives_dir = bids_layout.create_structure()
-
-            # Save dataset_description.json
-            bids_layout.save_dataset_description()
-
-            # Save participants.tsv
-            descriptions = [ds.description for ds in self.datasets]
-            bids_layout.save_participants(descriptions)
-
-            # Save BIDS sidecar files for each recording
-            self._save_bids_sidecar_files(bids_layout)
-
-            # Convert dataset to Zarr format inside derivatives
+            # Convert dataset to Zarr format
             log.info("Converting dataset to Zarr format...")
-            dataset_path = derivatives_dir / "dataset.zarr"
-
+            dataset_path = tmp_path / "dataset.zarr"
             self._convert_to_zarr_inline(
                 dataset_path,
                 compression,
                 compression_level,
             )
 
-            # Save dataset metadata (README.md)
+            # Save dataset metadata
             self._save_dataset_card(tmp_path)
 
             # Save format info
             format_info_path = tmp_path / "format_info.json"
-            with open(format_info_path, "w", encoding="utf-8") as f:
+            with open(format_info_path, "w") as f:
                 format_info = self._get_format_info_inline()
                 json.dump(
                     {
                         "format": "zarr",
-                        "bids_compatible": True,
-                        "pipeline_name": pipeline_name,
                         "compression": compression,
                         "compression_level": compression_level,
                         "braindecode_version": braindecode.__version__,
@@ -215,8 +185,8 @@ class HubDatasetMixin:
             # Default commit message
             if commit_message is None:
                 commit_message = (
-                    f"Upload EEG dataset in BIDS-like "
-                    f"Zarr format ({len(self.datasets)} recordings)"
+                    f"Upload EEG dataset in Zarr format "
+                    f"({len(self.datasets)} recordings)"
                 )
 
             # Upload folder to Hub
@@ -236,7 +206,7 @@ class HubDatasetMixin:
             except Exception as e:
                 raise RuntimeError(f"Failed to upload dataset: {e}")
 
-    def _save_dataset_card(self, path: Path, bids_compatible: bool = True) -> None:
+    def _save_dataset_card(self, path: Path) -> None:
         """Generate and save a dataset card (README.md) with metadata.
 
         Parameters
@@ -278,71 +248,13 @@ class HubDatasetMixin:
             sfreq=sfreq,
             data_type=data_type,
             n_windows=n_windows,
-            bids_compatible=True,
         )
 
         # Save README
         readme_path = path / "README.md"
-        with open(readme_path, "w", encoding="utf-8") as f:
+        with open(readme_path, "w") as f:
             f.write(readme_content)
 
-    def _save_bids_sidecar_files(
-        self, bids_layout: "bids_format.BIDSDerivativesLayout"
-    ) -> None:
-        """Save BIDS-compliant sidecar files for each recording.
-
-        This creates events.tsv, channels.tsv, and EEG sidecar JSON files
-        for each recording in a BIDS-like directory structure.
-
-        Parameters
-        ----------
-        bids_layout : BIDSDerivativesLayout
-            BIDS layout object for path generation.
-        """
-        dataset_type = get_dataset_type(self.datasets[0])
-
-        for i_ds, ds in enumerate(self.datasets):
-            # Get BIDS entities from description
-            description = ds.description if ds.description is not None else pd.Series()
-
-            # Get BIDSPath for this recording using mne_bids
-            bids_path = bids_layout.get_bids_path(description)
-
-            # Create subject directory
-            bids_path.mkdir(exist_ok=True)
-
-            # Get metadata and info based on dataset type
-            if dataset_type == "WindowsDataset":
-                info = ds.windows.info
-                metadata = ds.windows.metadata
-                sfreq = info["sfreq"]
-            elif dataset_type == "EEGWindowsDataset":
-                info = ds.raw.info
-                metadata = ds.metadata
-                sfreq = info["sfreq"]
-            elif dataset_type == "RawDataset":
-                info = ds.raw.info
-                metadata = None
-                sfreq = info["sfreq"]
-            else:
-                continue
-
-            # Determine task name from description or BIDSPath
-            task_name = bids_path.task or "unknown"
-
-            # Save BIDS sidecar files using mne_bids BIDSPath
-            bids_format.save_bids_sidecar_files(
-                bids_path=bids_path,
-                info=info,
-                metadata=metadata,
-                sfreq=sfreq,
-                task_name=str(task_name),
-            )
-
-            log.debug(
-                f"Saved BIDS sidecar files for recording {i_ds} to {bids_path.directory}"
-            )
-
     @classmethod
     def pull_from_hub(
         cls,
@@ -428,22 +340,8 @@ class HubDatasetMixin:
             else:
                 format_info = {}
 
-            # Determine if BIDS-compatible structure
-            bids_compatible = format_info.get("bids_compatible", False)
-            pipeline_name = format_info.get("pipeline_name", "braindecode")
-
-            # Find zarr dataset path
-            if bids_compatible:
-                # Look in derivatives folder
-                zarr_path = (
-                    Path(dataset_dir) / "derivatives" / pipeline_name / "dataset.zarr"
-                )
-                if not zarr_path.exists():
-                    # Fall back to root level (backwards compatibility)
-                    zarr_path = Path(dataset_dir) / "dataset.zarr"
-            else:
-                zarr_path = Path(dataset_dir) / "dataset.zarr"
-
+            # Load zarr dataset
+            zarr_path = Path(dataset_dir) / "dataset.zarr"
             if not zarr_path.exists():
                 raise FileNotFoundError(
                     f"Zarr dataset not found at {zarr_path}. "
@@ -452,17 +350,11 @@ class HubDatasetMixin:
 
             dataset = cls._load_from_zarr_inline(zarr_path, preload)
 
-            # Load BIDS metadata if available
-            if bids_compatible:
-                cls._load_bids_metadata(dataset, Path(dataset_dir), pipeline_name)
-
             log.info(f"Dataset loaded successfully from {repo_id}")
             log.info(f"Recordings: {len(dataset.datasets)}")
             log.info(
                 f"Total windows/samples: {format_info.get('total_samples', 'N/A')}"
             )
-            if bids_compatible:
-                log.info("BIDS sidecar files available in derivatives folder")
 
             return dataset
 
@@ -477,71 +369,6 @@ class HubDatasetMixin:
         except Exception as e:
             raise RuntimeError(f"Failed to load dataset from Hub: {e}")
 
-    @classmethod
-    def _load_bids_metadata(
-        cls,
-        dataset,
-        dataset_dir: Path,
-        pipeline_name: str,
-    ) -> None:
-        """Load BIDS metadata from sidecar files and attach to dataset.
-
-        Parameters
-        ----------
-        dataset : BaseConcatDataset
-            The loaded dataset to attach metadata to.
-        dataset_dir : Path
-            Root directory of the downloaded dataset.
-        pipeline_name : str
-            Name of the processing pipeline.
-        """
-        derivatives_dir = dataset_dir / "derivatives" / pipeline_name
-
-        # Load participants.tsv if available
-        participants_path = derivatives_dir / "participants.tsv"
-        if participants_path.exists():
-            try:
-                participants_df = pd.read_csv(participants_path, sep="\t")
-                # Store as attribute on the concat dataset
-                dataset.participants = participants_df
-                log.debug(
-                    f"Loaded participants info for {len(participants_df)} subjects"
-                )
-            except Exception as e:
-                log.warning(f"Failed to load participants.tsv: {e}")
-
-        # Create layout for path generation
-        bids_layout = bids_format.BIDSDerivativesLayout(
-            dataset_dir, pipeline_name=pipeline_name
-        )
-
-        # Try to load events.tsv files and attach to individual datasets
-        for i_ds, ds in enumerate(dataset.datasets):
-            description = ds.description if ds.description is not None else pd.Series()
-
-            # Get BIDSPath for this recording
-            bids_path = bids_layout.get_bids_path(description)
-
-            # Load events.tsv if available
-            events_path = bids_path.copy().update(suffix="events", extension=".tsv")
-            if events_path.fpath.exists():
-                try:
-                    events_df = pd.read_csv(events_path.fpath, sep="\t")
-                    ds.bids_events = events_df
-                    log.debug(f"Loaded events for recording {i_ds}")
-                except Exception as e:
-                    log.warning(f"Failed to load events for recording {i_ds}: {e}")
-
-            # Load channels.tsv if available
-            channels_path = bids_path.copy().update(suffix="channels", extension=".tsv")
-            if channels_path.fpath.exists():
-                try:
-                    channels_df = pd.read_csv(channels_path.fpath, sep="\t")
-                    ds.bids_channels = channels_df
-                    log.debug(f"Loaded channels for recording {i_ds}")
-                except Exception as e:
-                    log.warning(f"Failed to load channels for recording {i_ds}: {e}")
-
     def _convert_to_zarr_inline(
         self,
         output_path: Path,
@@ -1036,7 +863,6 @@ def _generate_readme_content(
     data_type: str,
     n_windows: int,
     format: str = "zarr",
-    bids_compatible: bool = True,
 ):
     """Generate README.md content for a dataset uploaded to the Hub."""
     # Use safe access for total size and format sfreq nicely
@@ -1045,52 +871,12 @@ def _generate_readme_content(
     )
     sfreq_str = f"{sfreq:g}" if sfreq is not None else "N/A"
 
-    bids_section = ""
-    if bids_compatible:
-        bids_section = """
-## BIDS-like Structure
-
-This dataset follows a BIDS derivatives-like structure for compatibility with
-neuroimaging tools while maintaining efficiency for deep learning:
-
-```
-derivatives/braindecode/
-├── dataset_description.json    # BIDS dataset description
-├── participants.tsv            # Subject-level metadata
-├── dataset.zarr/               # Main data (optimized for training)
-└── sub-<label>/
-    └── eeg/
-        ├── *_events.tsv        # Trial/window events
-        ├── *_channels.tsv      # Channel information
-        └── *_eeg.json          # Recording metadata
-```
-
-### Accessing BIDS Metadata
-
-After loading the dataset, BIDS metadata is available:
-
-```python
-# Access participants info
-if hasattr(dataset, "participants"):
-    print(dataset.participants)
-
-# Access events for a recording
-if hasattr(dataset.datasets[0], "bids_events"):
-    print(dataset.datasets[0].bids_events)
-
-# Access channel info
-if hasattr(dataset.datasets[0], "bids_channels"):
-    print(dataset.datasets[0].bids_channels)
-```
-"""
-
     return f"""---
 tags:
 - braindecode
 - eeg
 - neuroscience
 - brain-computer-interface
-- bids
 license: unknown
 ---
 
@@ -1109,7 +895,6 @@ This dataset was created using [braindecode](https://braindecode.org), a library
 | Number of windows / samples | {n_windows} |
 | Total size | {total_size_mb:.2f} MB |
 | Storage format | {format} |
-| BIDS compatible | {"Yes" if bids_compatible else "No"} |
 
 ## Usage
 
@@ -1148,7 +933,7 @@ To load this dataset::
         # y shape: [batch_size]
         # metainfo shape: [batch_size, 2] (start and end indices)
         # Process your batch...
-{bids_section}
+
 ## Dataset Format
 
 This dataset is stored in **Zarr** format, optimized for:

{braindecode-1.3.0.dev175563314 → braindecode-1.3.0.dev175887834}/braindecode/datasets/nmt.py

@@ -150,10 +150,7 @@ class NMT(BaseConcatDataset):
             os.path.join(path, "Labels.csv"), index_col="recordname"
         )
         if recording_ids is not None:
-            # Match metadata by record name instead of position to fix alignment bug
-            # when CSV order differs from sorted file order
-            selected_recordnames = [os.path.basename(fp) for fp in file_paths]
-            description = description.loc[selected_recordnames]
+            description = description.iloc[recording_ids]
         description.replace(
             {
                 "not specified": "X",
@@ -199,23 +196,18 @@ def _get_header(*args):
 
 def _fake_pd_read_csv(*args, **kwargs):
     # Create a list of lists to hold the data
-    # Updated to match the file IDs from _NMT_PATHS (0000036-0000042)
-    # to align with the mocked glob.glob return value
     data = [
-        ["0000036.edf", "normal", 35, "male", "train"],
-        ["0000037.edf", "abnormal", 28, "female", "test"],
-        ["0000038.edf", "normal", 62, "male", "train"],
-        ["0000039.edf", "abnormal", 41, "female", "test"],
-        ["0000040.edf", "normal", 19, "male", "train"],
-        ["0000041.edf", "abnormal", 55, "female", "test"],
-        ["0000042.edf", "normal", 71, "male", "train"],
+        ["0000001.edf", "normal", 35, "male", "train"],
+        ["0000002.edf", "abnormal", 28, "female", "test"],
+        ["0000003.edf", "normal", 62, "male", "train"],
+        ["0000004.edf", "abnormal", 41, "female", "test"],
+        ["0000005.edf", "normal", 19, "male", "train"],
+        ["0000006.edf", "abnormal", 55, "female", "test"],
+        ["0000007.edf", "normal", 71, "male", "train"],
     ]
 
     # Create the DataFrame, specifying column names
     df = pd.DataFrame(data, columns=["recordname", "label", "age", "gender", "loc"])
-    
-    # Set recordname as index to match the real pd.read_csv behavior with index_col="recordname"
-    df.set_index("recordname", inplace=True)
 
     return df
 

{braindecode-1.3.0.dev175563314 → braindecode-1.3.0.dev175887834}/braindecode/models/__init__.py

@@ -32,7 +32,6 @@ from .luna import LUNA
 from .medformer import MEDFormer
 from .msvtnet import MSVTNet
 from .patchedtransformer import PBT
-from .reve import REVE
 from .sccnet import SCCNet
 from .shallow_fbcsp import ShallowFBCSPNet
 from .signal_jepa import (
@@ -94,7 +93,6 @@ __all__ = [
     "MEDFormer",
     "MSVTNet",
     "PBT",
-    "REVE",
     "SCCNet",
     "ShallowFBCSPNet",
     "SignalJEPA",

{braindecode-1.3.0.dev175563314 → braindecode-1.3.0.dev175887834}/braindecode/models/base.py

@@ -192,7 +192,7 @@ class EEGModuleMixin(_BaseHubMixin, metaclass=NumpyDocstringInheritanceInitMeta)
             n_times is not None
             and input_window_seconds is not None
             and sfreq is not None
-            and n_times != round(input_window_seconds * sfreq)
+            and n_times != int(input_window_seconds * sfreq)
         ):
             raise ValueError(
                 f"{n_times=} different from {input_window_seconds=} * {sfreq=}"
@@ -236,7 +236,7 @@ class EEGModuleMixin(_BaseHubMixin, metaclass=NumpyDocstringInheritanceInitMeta)
             and self._input_window_seconds is not None
             and self._sfreq is not None
         ):
-            return round(self._input_window_seconds * self._sfreq)
+            return int(self._input_window_seconds * self._sfreq)
         elif self._n_times is None:
             raise ValueError(
                 "n_times could not be inferred. "

{braindecode-1.3.0.dev175563314 → braindecode-1.3.0.dev175887834}/braindecode/models/bendr.py

@@ -95,31 +95,6 @@ class BENDR(EEGModuleMixin, nn.Module):
         **self-supervised pre-training** on large, unlabeled EEG datasets (like TUEG) followed
         by subsequent fine-tuning on the specific downstream classification task [bendr]_.
 
-    .. important::
-       **Pre-trained Weights Available**
-
-       This model has pre-trained weights available on the Hugging Face Hub.
-       You can load them using:
-
-       .. code-block:: python
-
-           from braindecode.models import BENDR
-
-           # Load pre-trained model from Hugging Face Hub
-           # you can specify `n_outputs` for your downstream task
-           model = BENDR.from_pretrained("braindecode/braindecode-bendr", n_outputs=2)
-
-       To push your own trained model to the Hub:
-
-       .. code-block:: python
-
-           # After training your model
-           model.push_to_hub(
-               repo_id="username/my-bendr-model", commit_message="Upload trained BENDR model"
-           )
-
-       Requires installing ``braindecode[hug]`` for Hub integration.
-
     Parameters
     ----------
     encoder_h : int, default=512

{braindecode-1.3.0.dev175563314 → braindecode-1.3.0.dev175887834}/braindecode/models/biot.py

@@ -41,35 +41,6 @@ class BIOT(EEGModuleMixin, nn.Module):
     linear layer that takes the output of the `BIOTEncoder` and outputs
     the classification probabilities.
 
-    .. important::
-       **Pre-trained Weights Available**
-
-       This model has pre-trained weights available on the Hugging Face Hub.
-       You can load them using:
-
-       .. code-block:: python
-
-           from braindecode.models import BIOT
-
-           # Load the original pre-trained model from Hugging Face Hub
-           # For 16-channel models:
-           model = BIOT.from_pretrained("braindecode/biot-pretrained-prest-16chs")
-
-           # For 18-channel models:
-           model = BIOT.from_pretrained("braindecode/biot-pretrained-shhs-prest-18chs")
-           model = BIOT.from_pretrained("braindecode/biot-pretrained-six-datasets-18chs")
-
-       To push your own trained model to the Hub:
-
-       .. code-block:: python
-
-           # After training your model
-           model.push_to_hub(
-               repo_id="username/my-biot-model", commit_message="Upload trained BIOT model"
-           )
-
-       Requires installing ``braindecode[hug]`` for Hub integration.
-
     .. versionadded:: 0.9
 
     Parameters

{braindecode-1.3.0.dev175563314 → braindecode-1.3.0.dev175887834}/braindecode/models/config.py

@@ -151,12 +151,12 @@ def make_model_config(
             and sfreq is not None
             and input_window_seconds is not None
         ):
-            if n_times != round(input_window_seconds * sfreq):
+            if n_times != int(input_window_seconds * sfreq):
                 raise ValueError(
                     f"Provided {n_times=} does not match {input_window_seconds=} * {sfreq=}."
                 )
         elif n_times is None and sfreq is not None and input_window_seconds is not None:
-            data["n_times"] = round(input_window_seconds * sfreq)
+            data["n_times"] = int(input_window_seconds * sfreq)
         elif sfreq is None and n_times is not None and input_window_seconds is not None:
             data["sfreq"] = n_times / input_window_seconds
         elif input_window_seconds is None and n_times is not None and sfreq is not None: