pymaid.ClustResults

class pymaid.ClustResults(mat, labels=None, mat_type='distance')

Class to handle, analyze and plot similarity/distance matrices.

Contains thin wrappers for scipy.cluster.

dist_mat

Type:

Distance matrix (0=similar, 1=dissimilar)

sim_mat

Type:

Similarity matrix (0=dissimilar, 1=similar)

linkage

to generate linkage. By default, WARD’s algorithm is used.

Type:

Hierarchical clustering. Run pymaid.ClustResults.cluster()

leafs

Type:

list of skids

Examples

>>> import matplotlib.pyplot as plt
>>> # Get a bunch of neurons
>>> nl = pymaid.get_neuron('annotation:glomerulus DA1')
>>> # Perform connectivity clustering
>>> res = pymaid.cluster_by_connectivity(nl)
>>> # `res` is a ClustResults object with handy methods:
>>> res.plot_matrix()
>>> plt.show()
>>> # Extract 5 clusters
>>> res.get_clusters(5, criterion = 'maxclust')

Initialize class instance.

Parameters:

• mat (numpy.array | pandas.DataFrame) – Distance or similarity matrix.

• labels (list, optional) – Labels for matrix.

• mat_type ('distance' | 'similarity', default = 'distance') –

  Sets the type of input matrix:

  • ’similarity’ = high values are more similar

  • ’distance’ = low values are more similar

  The “missing” matrix type will be computed. For clustering, plotting, etc. distance matrices are used.

__init__(mat, labels=None, mat_type='distance')

Initialize class instance.

Parameters:

• mat (numpy.array | pandas.DataFrame) – Distance or similarity matrix.

• labels (list, optional) – Labels for matrix.

• mat_type ('distance' | 'similarity', default = 'distance') –

  Sets the type of input matrix:

  • ’similarity’ = high values are more similar

  • ’distance’ = low values are more similar

  The “missing” matrix type will be computed. For clustering, plotting, etc. distance matrices are used.

Methods

__init__(mat[, labels, mat_type])	Initialize class instance.
calc_agg_coeff()	Return the agglomerative coefficient.
calc_cophenet()	Return Cophenetic Correlation coefficient of your clustering.
cluster([method])	Cluster distance matrix.
get_clusters(k[, criterion, return_type])	Get get clusters.
get_colormap([k, criterion])	Generate colormap based on clustering.
get_leafs([use_labels])	Retrieve leaf labels.
plot3d([k, criterion])	Plot neuron using pymaid.plot.plot3d().
plot_dendrogram([color_threshold, ...])	Plot dendrogram using matplotlib.
plot_matrix()	Plot distance matrix and dendrogram using matplotlib.
plot_matrix2(**kwargs)	Plot distance matrix and dendrogram using seaborn.
to_selection([fname, k, criterion])	Convert clustered neurons into json file.
to_tree()	Turn linkage to ete3 tree.