immuneML.ml_methods.clustering package

Submodules

immuneML.ml_methods.clustering.AgglomerativeClustering module

class immuneML.ml_methods.clustering.AgglomerativeClustering.AgglomerativeClustering(name=None, **kwargs)

Bases: ClusteringMethod

Agglomerative clustering method which wraps scikit-learn’s clustering of the same name. Input arguments for the method are the same as supported by scikit-learn (see AgglomerativeClustering scikit-learn documentation for details).

YAML specification:

definitions:
    ml_methods:
        my_agglomerative_clustering:
            AgglomerativeClustering:
                # arguments as defined by scikit-learn
                n_clusters: 3
                linkage: 'ward'

fit(dataset: Dataset)

fit_predict(dataset: Dataset)

predict(dataset: Dataset)

transform(dataset: Dataset)

immuneML.ml_methods.clustering.ClusteringMethod module

class immuneML.ml_methods.clustering.ClusteringMethod.ClusteringMethod(name: str = None)

Bases: object

Clustering methods are algorithms which can be used to cluster repertoires, receptors or sequences without using external label information (such as disease or antigen binding state)

These methods can be used in the Clustering instruction.

DOCS_TITLE = 'Clustering methods'

abstract fit(dataset: Dataset)

abstract fit_predict(dataset: Dataset)

abstract predict(dataset: Dataset)

abstract transform(dataset: Dataset)

immuneML.ml_methods.clustering.ClusteringMethod.get_data_for_clustering(dataset: Dataset)

immuneML.ml_methods.clustering.DBSCAN module

class immuneML.ml_methods.clustering.DBSCAN.DBSCAN(name=None, **kwargs)

Bases: ClusteringMethod

DBSCAN method which wraps scikit-learn’s clustering of the same name. Input arguments for the method are the same as supported by scikit-learn (see DBSCAN scikit-learn documentation for details).

YAML specification:

definitions:
    ml_methods:
        my_dbscan:
            DBSCAN:
                # arguments as defined by scikit-learn
                eps: 0.5
                min_samples: 5

fit(dataset: Dataset)

fit_predict(dataset: Dataset)

predict(dataset: Dataset)

transform(dataset: Dataset)

immuneML.ml_methods.clustering.KMeans module

class immuneML.ml_methods.clustering.KMeans.KMeans(name=None, **kwargs)

Bases: ClusteringMethod

k-means clustering method which wraps scikit-learn’s KMeans. Input arguments for the method are the same as supported by scikit-learn (see KMeans scikit-learn documentation for details).

YAML specification:

definitions:
    ml_methods:
        my_kmeans:
            KMeans:
                # arguments as defined by scikit-learn
                n_clusters: 2

fit(dataset: Dataset)

fit_predict(dataset: Dataset)

predict(dataset: Dataset)

transform(dataset: Dataset)

Module contents