timeseries-clustering.py

from tslearn.clustering import TimeSeriesKMeans
from tslearn.utils import to_time_series_dataset
import pandas as pd

data = [
    [0, 1, 2, 3, 4],
    [1, 2, 3, 4, 0],
    [0, 1, 2, 3, 4],
    [0, 1, 2, 3, 4],
    [5, 6, 3, 4, 0],
    [5, 6, 3, 4, 0, 5, 6, 3, 4, 0],
]
X = to_time_series_dataset(data)

km = TimeSeriesKMeans(n_clusters=4, metric='dtw', verbose=False, random_state=0, max_iter=5, n_init=3)
labels = km.fit_predict(X)

# Using pandas DataFrame
df = pd.DataFrame([
    {'id': i, 'timeseries': x}
    for i, x in enumerate(data)
]).set_index('turnaround_id')

X = to_time_series_dataset(df['timeseries'].to_list())

km = TimeSeriesKMeans(n_clusters=4, metric='dtw', verbose=False, random_state=0, max_iter=5, n_init=3)
df['labels'] = km.fit_predict(X)

# Silhouette
from tslearn.clustering import silhouette_score

silhouette_score(X, labels)