An interactive EDA session¶

Stub. This notebook is part of the documentation scaffold. The final version will mirror bench/session_vs_oneshot.py — opening a persistent session, running several follow-up queries against the same catalogs, and materialize()-ing an intermediate crossmatch.

Planned flow:

1. with acid.connect(catalogs, workers=8) as s: open a session against the catalogs from Your first crossmatch.
2. Run three follow-up queries from a warm pool; time the difference against an equivalent loop of acid.sql(...) calls.
3. s.materialize("nearby", ...) to cache a heavy 3-way XMATCH; then iterate on filters and aggregates against nearby — subsecond per query.
4. Switch a result to Polars (r.to_polars()) and do a quick group-by-magnitude analysis.

See Sessions and materialize() for the concept; this notebook will be the worked example.