# Results

Evaluates three pipelines (CSP+LDA, TS+SVM, MDM) across five motor imagery datasets with
MOABB's `CrossSubjectEvaluation`, then runs both a frequentist (`nma`) and a Bayesian
(`bnma`) network meta-analysis on the resulting evaluation results.

```python
from moabb.datasets import BNCI2014_001, BNCI2014_004, Cho2017, Lee2019_MI, PhysionetMI
from moabb.evaluations import CrossSubjectEvaluation
from moabb.paradigms import LeftRightImagery
from moabbr import nma, bnma
from pyriemann.classification import MDM
from pyriemann.estimation import Covariances
from pyriemann.spatialfilters import CSP
from pyriemann.tangentspace import TangentSpace
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis as LDA
from sklearn.pipeline import make_pipeline
from sklearn.svm import SVC

pipelines = {
    "CSP+LDA": make_pipeline(
        Covariances(estimator="oas"),
        CSP(nfilter=6),
        LDA(solver="svd"),
    ),
    "TS+SVM": make_pipeline(
        Covariances(estimator="oas"),
        TangentSpace(metric="riemann"),
        SVC(kernel="linear"),
    ),
    "MDM": make_pipeline(Covariances(estimator="oas"), MDM(metric="riemann")),
}

datasets = [BNCI2014_001(), BNCI2014_004(), Cho2017(), Lee2019_MI(), PhysionetMI()]
paradigm = LeftRightImagery(resample=128)
evaluation = CrossSubjectEvaluation(
    paradigm=paradigm,
    datasets=datasets,
    overwrite=False,
    n_splits=min(dataset.metadata.participants.n_subjects for dataset in datasets),
    cache_config=dict(
        use=True,
        save_array=True,
        overwrite_array=False,
    ),
)

results = evaluation.process(pipelines)
freq = nma(results)
bayes = bnma(results)
```