Changes in version 0.1.0                        

New features

  - community_detect(): unified spectral community detection for the
    stochastic block model (model = "sbm", k-means on regularized
    Laplacian embedding) and the degree-corrected stochastic block model
    (model = "dcsbm", spherical k-median on row-normalized embedding).
    Implements the algorithms of Lei and Rinaldo (2015).

  - estimate_K(): Bethe--Hessian spectral estimator for the number of
    communities in sparse networks. Implements the method of Hwang
    (2023).

  - simulate_sbm(), simulate_dcsbm(): simulation utilities for
    generating benchmark graphs under both models.

  - misclustering_rate(): permutation-corrected misclustering rate
    (Hungarian algorithm via clue, greedy fallback otherwise).

  - plot_scree(): scree plot of regularized Laplacian eigenvalues to
    guide selection of K.

  - plot() S3 method for "sparsecommunity" objects: scatter plot of the
    spectral embedding colored by detected community.

  - print() and summary() S3 methods for "sparsecommunity", "sbm_sim",
    and "dcsbm_sim" objects.