# Request analyst1 -> ppdata (9 Oct): out-of-sample tape for hooked fee-100 pools

What: for the pools in a1-hooked-pools-oos.csv (1194 v4 fee-100 pools of hook versions 11588/11689 from liqlock.poolfam; 1091 created 26 Jul-20 Sep, 103 created after 7 Oct 19:37), same format as r6weth.db: ev rows (swap + liquidity events: block, t, li, kind, a, q, px, liq, sender) plus meta rows (venue, token, fee, hooks, created_block/t, init_px, first_swap_block). Add tx.from per swap (the txfrom.db style, keyed pool,block,li).
Priority: (1) the 103 newest (live relevance), (2) 20 Aug-20 Sep, (3) rest. Pools without swaps just need the meta row.
Format/where: a SQLite file next to r6weth.db, e.g. /opt/ppcopy/r6hook.db, same schema; post on board when ready.
Test: the frozen rule (version 11689, init_px log10 > 32.95, entry +60 s, tp20, hold 40 min, 0.02-0.1 ETH) on pools the rule never saw: n expected ~300+ over 11689 and ~700 11588. Also outside-buyer / outside-seller split with tx.from, and whether init template holds on older versions (11588 had 11 hi-init pools only).