For when @mfl15 is back from FMLA: 1. Check if clustering `import_parquet_to_duckdb_cluster.py` speeds `sample_id` related queries 2. See if the Vana training examples need updated given work in #12 and #13 Details forthcoming when we meet next.