You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The precalculated app was refactored in #24 to separate dim reduction from KMeans so users can project once and run multiple k values without re-projecting. The embed_explore app still couples them in a single "Run Clustering" action.
We'd like to mirror the precalculated app's split:
Replace render_clustering_section() in apps/embed_explore/components/sidebar.py with separate render_projection_section() and render_kmeans_section()
Replace hardcoded color_col = 'cluster' in shared/components/visualization.py:127 with the "Color by" dropdown pattern. Color by different KMeans cluster result
Preserve KMeans columns on re-projection with PCA, TSNE, UMAP
The
precalculatedapp was refactored in #24 to separate dim reduction from KMeans so users can project once and run multiplekvalues without re-projecting. Theembed_exploreapp still couples them in a single "Run Clustering" action.We'd like to mirror the precalculated app's split:
render_clustering_section()inapps/embed_explore/components/sidebar.pywith separaterender_projection_section()andrender_kmeans_section()ClusteringService.run_dim_reduction_safe()andrun_kmeans_only_safe()color_col = 'cluster'inshared/components/visualization.py:127with the "Color by" dropdown pattern. Color by different KMeans cluster result