Iteration class

class tclust._iteration.Iteration(object)[source]

Helper class for TClust package.

obj

Value of the objective function of the current solution.

Type

float

labels_

Cluster assignment for each observation. Shape=(nobs, ).

Type

ndarray

csize

Number of observations that belong to each cluster. Shape=(1, k).

Type

ndarray

cw

Weights of each cluster. Shape=(k, )

Type

ndarray

sigma

Covariance matrices of each cluster. Shape=(ndim, ndim, k).

Type

ndarray

center

Centers of each cluster (column-wise). Shape=(k, ndim).

Type

ndarray

code

Return code (supplied by functions like findClusterLabels).

Type

int

z_ij

Cluster assignment for each sample. Shape=(nobs, k).

Type

ndarray

lambd

Diagonal values for tkmeans. Shape=(k, ).

Type

ndarray

fill(nobs, ndim, k)[source]
Parameters
  • nobs (int) – Number of observations in the sample.

  • ndim (int) – Number of dimensions.

  • k (int) – Number of clusters

update(obj=None, labels_=None, csize=None, cw=None, sigma=None, center=None, z_ij=None, lambd=None, code=None)[source]

Updates the values of the Iteration object.

Parameters
  • obj – value of the objective function

  • labels – cluster assignment

  • csize – cluster sizes

  • cw – cluster weights

  • sigma – covariance matrices of each cluster

  • center – centers of each cluster (column-wise)

  • z_ij – cluster assignment

  • lambd – diagonal values for tkmeans

  • code – return code

Returns

self