kmeans.train(clusters=n_clusters, max_iter=m_iter)
kmeans.predict()
metric_collection.append(kmeans.metrics)
test = kmeans.get_clustered
if pairplot_flag:
try:
vis.make_pairplot(
kmeans.get_clustered, dp.get_ages,
f'{train_l}_trainL_{n_clusters}_clusters')
print(
f'pairplot for train distribution {train_l} clusters {n_clusters} built successfully'
)
except Exception as e:
print(f'{e} has been dropped')
analyzer.metric_collection('KMeans', metric_collection)
analyzer.best_clustering_find()
else:
kmeans_best = KMeansClassifier(dp.dataset_no_useless, 15000)
kmeans_best.train(3, 900)
kmeans_best.predict()
kmeans_best.clusters_to_excel()
analyzer.separate_clusters(kmeans_best.get_clustered)
vis.make_pairplot(kmeans_best.get_clustered, dp.get_ages,
f'best_clustering_pairplot')
forest_test_scores = []
forest_train_scores = []
train_l_list = [
i
for i in range(int(len(kmeans_best.get_clustered) *
0.2), int(len(kmeans_best.get_clustered) *
0.8), 100)
