moabb.datasets.Ma2022#
- class moabb.datasets.Ma2022(subjects=None, sessions=None, *, return_all_modalities=False)[source]#
Bases:
BaseBIDSDataset[source]Dataset Snapshot
Ma2022
Imagery, 2 classes (left_hand vs right_hand)
Class Labels: left_hand, right_hand
Citation & Impact
- Paper DOI10.1038/s41597-022-01647-1
- CitationsLoading…
- Public APICrossref | OpenAlex
Stimulus Protocol
4s task window per trial · 2-class imagery paradigm · 1 runs/session across 5 sessions
HED Event TagsHED tagsSource: MOABB BIDS HED annotation mapping.
left_handSensory-eventAgent-actionright_handSensory-eventAgent-actionHED tree view
Tree · left_hand
├─ Sensory-event │ ├─ Experimental-stimulus │ └─ Visual-presentation └─ Agent-action └─ Imagine ├─ Move └─ Left └─ HandTree · right_hand
├─ Sensory-event │ ├─ Experimental-stimulus │ └─ Visual-presentation └─ Agent-action └─ Imagine ├─ Move └─ Right └─ HandChannel SummaryTotal channels32EEG32 (Ag/AgCl)Montagestandard_1020Sampling250 HzReferenceM1 (unipolar)Notch / line50 HzThis diagram is automatically generated from MOABB metadata. Please consult the original publication to confirm the experimental protocol details.
Cross-session motor imagery dataset (SHU) from Ma et al. 2022.
Dataset from [1].
The SHU dataset contains EEG recordings from 25 healthy, BCI-naive subjects (13 males, 12 females, aged 20-24 years) performing cued left- vs right-hand grasping motor imagery. Each subject completed five independent sessions recorded on five different days, 2 to 3 days apart, which makes the dataset specifically suited for studying cross-session variability in motor imagery BCIs.
Each session was designed with 100 trials (50 left-hand, 50 right-hand, randomized order); the data paper’s Table 2 reports “90 to 100” trials per session. The released files retain 74 to 100 trials per session after the source-side bad-segment rejection described in the data paper, for 11,988 trials in total. Signals were recorded from 32 EEG channels (10-10 according to the paper, called 10-20 in the source sidecar; unipolar reference on M1, ground on AFz) at 250 Hz. Only the 4 s motor imagery window is stored (1000 samples per trial), so the analysis interval spans the full stored window. The source sidecar describes an 8 s protocol whereas the paper text states 7.5 s; missing preparation and cue segments are not reconstructed.
Important
The released data is preprocessed: bad segments were removed, the baseline was corrected, and a 0.5-40 Hz FIR band-pass filter was applied by the authors before disclosure.
This loader reads the authors’ EDF release, hosted in BIDS form on NEMAR (
nm000288). Each EDF contains the retained 4 s windows concatenated in time, not continuous amplifier recordings. BIDSDatasetType: rawdescribes the deposit layout, not its processing state. Events and bad-channel flags are read from the BIDS sidecars; MNE converts the EDF microvolt calibration to SI volts. Zero-durationEDGE boundaryannotations mark the joins between stored windows, so MNE’s default filtering does not mix neighboring, discontinuous trials.Nine sessions contain a channel numerically near zero in the authors’ MATLAB release, with blank EDF calibration fields. The deposit assigns a small replacement physical range and marks the channel bad; this is not a recovery of the missing calibration. The untouched EDF is preserved under
sourcedata/. Bad channels remain flagged, without interpolation or deletion. Since MOABB’s default EEG selection excludes bads, full-dataset analyses must use a common good-channel set (exclude F3, T6 and A2). Explicitly picking a bad channel includes its replacement-calibrated signal. Channel positions are template estimates fromstandard_1020(calledcolin27_1020in newer MNE versions), not measured locations.Note
NEMAR is the only download source for this EDF/BIDS loader, including when the provider is set to
upstream. Failures propagate: there is no fallback to the scientifically different MATLAB representation.data_pathreturns the five EDF paths; sessions are numbered"0"to"4"in MOABB. The codeMa-edf2022isolates downloads, caches and evaluation results from the former MATLAB loader.This dataset is from the same laboratory as
Yang2025(WBCIC-SHU, a distinct 2025 multi-day recording) and is unrelated toMa2020(different team, different recording).References
[1]J. Ma, B. Yang, W. Qiu, Y. Li, S. Gao, and X. Xia, “A large EEG dataset for studying cross-session variability in motor imagery brain-computer interface,” Scientific Data, vol. 9, p. 531, 2022. DOI: 10.1038/s41597-022-01647-1
from moabb.datasets import Ma2022 dataset = Ma2022() data = dataset.get_data(subjects=[1]) print(data[1])
Dataset summary
#Subj
25
#Chan
32
#Classes
2
#Trials / class
239.76
Trials length
4 s
Freq
250 Hz
#Sessions
5
#Runs
1
Total_trials
11988
Participants
Population: healthy
Age: 22.52 (range: 20-24) years
BCI experience: naive
Equipment
Amplifier: Wuhan Greentech 32-channel Ag/AgCl cap, Brickcom wireless amplifier
Electrodes: Ag/AgCl
Montage: standard_1020
Reference: M1 (unipolar)
Preprocessing
Data state: preprocessed
Bandpass filter: 0.5-40 Hz
Steps: bad segment rejection (EEGLAB amplitude > 100 uV, manually confirmed), baseline removal, 0.5-40 Hz FIR band-pass filter, epoching to the 4 s motor imagery window
Notes: Preprocessing was applied by the data authors before disclosure; the released trials are the 4 s motor imagery segments only.
Data Access
DOI: 10.1038/s41597-022-01647-1
Data URL: https://nemar.org/dataexplorer/detail?dataset_id=nm000288
Repository: NEMAR
Experimental Protocol
Paradigm: imagery
Task type: motor_imagery_grasping
Feedback: none
Stimulus: visual cue
Notes
Added in version 1.8.0.
- __init__(subjects=None, sessions=None, *, return_all_modalities=False)[source]#
Initialize function for the BaseDataset.
- property all_subjects#
Full list of subjects available in this dataset (unfiltered).
- convert_to_bids(path=None, subjects=None, overwrite=False, format='EDF', verbose=None, generate_figures=False)[source]#
Convert the dataset to BIDS format.
Saves the raw EEG data in a BIDS-compliant directory structure. Unlike the caching mechanism (see
CacheConfig), the files produced here do not contain a processing-pipeline hash (desc-<hash>) in their names, making the output a clean, shareable BIDS dataset.- Parameters:
path (str |
Path| None) – Directory under which the BIDS dataset will be written. IfNonethe default MNE data directory is used (same default as the rest of MOABB).subjects (list of int | None) – Subject numbers to convert. If
None, all subjects insubject_listare converted.overwrite (bool) – If
True, existing BIDS files for a subject are removed before saving. Default isFalse.format (str) – The file format for the raw EEG data. Supported values are
"EDF"(default),"BrainVision", and"EEGLAB".verbose (str | None) – Verbosity level forwarded to MNE/MNE-BIDS.
generate_figures (bool) – If
True, generate interactive neural signature HTML figures in{bids_root}/derivatives/neural_signatures/. Requiresplotly(pip install moabb[interactive]). Default isFalse.
- Returns:
bids_root – Path to the root of the written BIDS dataset.
- Return type:
Examples
>>> from moabb.datasets import AlexMI >>> dataset = AlexMI() >>> bids_root = dataset.convert_to_bids(path="/tmp/bids", subjects=[1])
Notes
Use
CacheConfigto configure caching forget_data(). Usemoabb.datasets.bids_interface.get_bids_rootto get the BIDS root path.Added in version 1.5.
- data_path(subject, path=None, force_update=False, update_path=None, verbose=None)[source]#
Get path to local copy of a subject data.
- Parameters:
subject (int) – Number of subject to use
path (None | str) – Location of where to look for the data storing location. If None, the environment variable or config parameter
MNE_DATASETS_(dataset)_PATHis used. If it doesn’t exist, the “~/mne_data” directory is used. If the dataset is not found under the given path, the data will be automatically downloaded to the specified folder.force_update (bool) – Force update of the dataset even if a local copy exists.
update_path (bool | None Deprecated) – If True, set the MNE_DATASETS_(dataset)_PATH in mne-python config to the given path. If None, the user is prompted.
verbose (bool, str, int, or None) – If not None, override default verbose level (see
mne.verbose()).
- Returns:
path – Local path to the given data file. This path is contained inside a list of length one, for compatibility.
- Return type:
- download(subject_list=None, path=None, force_update=False, update_path=None, accept=False, verbose=None)[source]#
Download BIDS EDFs, not the original sourcedata distribution.
- get_additional_metadata(subject, session, run)[source]#
Load additional metadata for a specific subject, session, and run.
- Parameters:
- Returns:
A DataFrame containing the additional metadata if available, otherwise None.
- Return type:
None |
pandas.DataFrame
- get_block_repetition(paradigm, subjects, block_list, repetition_list)[source]#
Select data for all provided subjects, blocks and repetitions.
subject -> session -> run -> block -> repetition
See also
- get_data(subjects=None, cache_config=None, process_pipeline=None, n_jobs=1)[source]#
Return the data corresponding to a list of subjects.
The returned data is a dictionary with the following structure:
data = {"subject_id": {"session_id": {"run_id": run}}}
subjects are on top, then we have sessions, then runs. A sessions is a recording done in a single day, without removing the EEG cap. A session is constitued of at least one run. A run is a single contiguous recording. Some dataset break session in multiple runs.
Processing steps can optionally be applied to the data using the
*_pipelinearguments. These pipelines are applied in the following order:raw_pipeline->epochs_pipeline->array_pipeline. If a*_pipelineargument isNone, the step will be skipped. Therefore, thearray_pipelinemay either receive amne.io.Rawor amne.Epochsobject as input depending on whetherepochs_pipelineisNoneor not.- Parameters:
subjects (List of int) – List of subject number
cache_config (dict |
CacheConfig) – Configuration for caching of datasets. SeeCacheConfigfor details.process_pipeline (
sklearn.pipeline.Pipeline| None) – Optional processing pipeline to apply to the data. To generate an adequate pipeline, we recommend usingmoabb.make_process_pipelines(). This pipeline will receivemne.io.BaseRawobjects. The steps names of this pipeline should be elements ofStepType. According to their name, the steps should either return amne.io.BaseRaw, amne.Epochs, or anumpy.ndarray. This pipeline must be “fixed” because it will not be trained, i.e. no call tofitwill be made.n_jobs (int) – Number of jobs to run in parallel over subjects (passed to
joblib.Parallel). Default1(sequential). Per-subject processing (reading, filtering, resampling, epoching) is independent, so this gives a near-linear speedup for datasets with many subjects.
- Returns:
data – dict containing the raw data
- Return type:
Dict
- property metadata[source]#
Return structured metadata for this dataset.
Returns the DatasetMetadata object from the centralized catalog, or None if metadata is not available for this dataset.
- Returns:
The metadata object containing acquisition parameters, participant demographics, experiment details, and documentation. Returns None if no metadata is registered for this dataset.
- Return type:
DatasetMetadata| None
Examples
>>> from moabb.datasets import BNCI2014_001 >>> dataset = BNCI2014_001() >>> dataset.metadata.participants.n_subjects 9 >>> dataset.metadata.acquisition.sampling_rate 250.0
- sourcedata_path(subject=None, path=None, force_update=False, verbose=None)[source]#
Get the dataset’s original pre-BIDS distribution from NEMAR.
Where
data_path()fetches the original files from the upstream host, this fetches the copy NEMAR mirrors undersourcedata/. The files keep their upstream names, so the two are interchangeable in content – but NEMAR stays reachable when the upstream host is slow, rate-limited, behind a bot gate, or retired.- Parameters:
subject (int | str | None) – Restrict the download to one subject, resolved through the deposit’s
sourcedata_provenance.json. Deposits enriched before that manifest recorded subjects fall back to the whole tree with a warning. WhenNonethe whole tree is fetched.path (None | str) – Base path where MOABB stores datasets.
force_update (bool) – Re-fetch even when a local copy is present.
verbose (bool, str, int, or None) – If not None, override default verbose level.
- Returns:
Local path to the
sourcedatadirectory.- Return type:
- Raises:
ValueError – If the dataset declares no
nemar_id.moabb.datasets.download.NemarDownloadError – If the download fails or the deposit publishes no
sourcedata/.