kmeans error: __init__() got an unexpected keyword argument 'precompute_distances'

[ostwaldj]

Created a new virtual env with only Texthero and its dependencies installed. A year or more ago it worked, but now when I try to run kmeans, e.g.
.pipe(hero.kmeans,n_clusters=5)
it errors:
init() got an unexpected keyword argument 'precompute_distances'

This happens with my own data and code, and also when I paste the BBCSport sample code in as well.

[PhelaPoscam]

Same problem.

Seems like sklearn.cluster.KMeans dont use precompute_distances anymore

https://scikit-learn.org/stable/modules/generated/sklearn.cluster.KMeans.html?highlight=kmeans#sklearn.cluster.KMeans

You can comment the lines "precompute_distances="auto"" and "#n_jobs=n_jobs," in representation.py and will work

def kmeans(
    s: pd.Series,
    n_clusters=5,
    init="k-means++",
    n_init=10,
    max_iter=300,
    tol=0.0001,
    precompute_distances="auto",
    verbose=0,
    random_state=None,
    copy_x=True,
    n_jobs=-1,
    algorithm="auto",
):
    """
    Perform K-means clustering algorithm.
    """
    vectors = list(s)
    kmeans = KMeans(
        n_clusters=n_clusters,
        init=init,
        n_init=n_init,
        max_iter=max_iter,
        tol=tol,
        #precompute_distances=precompute_distances,
        verbose=verbose,
        random_state=random_state,
        copy_x=copy_x,
        #n_jobs=n_jobs,
        algorithm=algorithm,
    ).fit(vectors)
    return pd.Series(kmeans.predict(vectors), index=s.index)

[anu43]

it's still present. when could it be re-evaluated?

[jbesomi]

Thanks for letting me know. I will look into that in the next few days and release a new version to fix that issue.