tickstream — FREE PREVIEW: NQ ticks (2 months) + Level-2 depth (1 week)
=======================================================================

A free, unmodified excerpt of the paid "NQ Tick + L2 Archive" at
https://tick-stream.xyz/historical-data — the exact same data the API serves.

CONTENTS
  nq-ticks-2025-02.parquet     every NQ trade, Feb 2025   (10.2M rows)
  nq-ticks-2025-03.parquet     every NQ trade, Mar 2025   (12.7M rows)
  nq-l2-book-2025-03-03.parquet, nq-l2-book-2025-03-04.parquet
                               FULL Level-2 order-book depth, two complete
                               trading days (Mar 3-4 2025), 34,692,616 rows

TICKS schema (one row per trade):
  ts      int64   event time, epoch MICROSECONDS (UTC)
  price   double  trade price
  size    int64   contracts
  side    int32   aggressor: 1 = buy (at ask), -1 = sell (at bid), 0 = unknown

LEVEL-2 schema (one row per book level update):
  ts      int64   event time, epoch MICROSECONDS (UTC)
  side    int32   1 = bid, 0 = ask
  level   int32   book level (1 = best)
  price   double  price at that level
  size    double  resting size at that level
  flag    int32   0 = snapshot row

Read it (Python):
  import pandas as pd
  book = pd.read_parquet("nq-l2-book-2025-03-03.parquet")
  book["time"] = pd.to_datetime(book["ts"], unit="us", utc=True)

Notes:
- Full 23h Globex sessions, front-month NQ.
- Trade aggressor side is quote-rule inferred against the prevailing BBO for this
  period (matches the paid archive; from March 2026 the live archive carries the
  exchange-reported aggressor).
- The paid archive has this L2 depth for EVERY day 2019 -> live via /v1/history/book.
- License: free for personal research and backtesting. Redistribution or resale
  is not permitted. Data provided as-is, no warranty. Not financial advice.

Full archive (2019 -> live: ticks, Level-2, options chains):
https://tick-stream.xyz/historical-data
Questions: support@tick-stream.xyz
