ForexBestRobots · Tick volume guide · V219

Purpose: reproduce quote-count measurements, not trading performance.
Source: HistData Generic ASCII EURUSD tick archive, January 2024.
Source timezone: fixed EST, UTC−05:00; add five hours to obtain UTC.

N counts every source quote record. Q counts the first archive record once,
then only records where Bid or Ask changes from the preceding archive record.
The Q comparison is not restarted at candle boundaries. Both are declared
reconstructions, not exports of a broker's native MT4 volume.

The source Volume column contains only zero. It is not used as the counter.
Neither count measures executed lots or market-wide traded currency units.
Each spread is Ask minus Bid; its median is per record, not time-weighted.

CSV prices and spreads use integer millionths; EURUSD one pip is 100 units.
start_utc labels [start,end) intervals. Empty price intervals are not filled.
m is the number of distinct observed minute intervals inside the group.
A complete M5 group has m=5. Earlier reference medians require every one
of the immediately preceding L chronological intervals to be present and
complete, excluding the current candle. The viewer also requires a complete
current group and a positive median before displaying a ratio.

The interactive example is 2024-01-08 07:00–17:00 UTC. Earlier rows are retained
for reference windows. The activity, narrow-range and spread examples were
selected retrospectively; their exact rules and timestamps are in data-summary.

The original monthly archive is not redistributed in this public package.
source-manifest.json identifies the original page, format, size and SHA-256.
The included tables can be inspected directly without running any software.

For optional full reproduction, use Python 3 with pandas, numpy and lxml.
fetch_ticks.py uses only the source's displayed free historical-data form.
It creates source/ and research/, downloads the original monthly ZIP and
records its current SHA-256. Compare it with the supplied source manifest
before replacing any reference data; stop if the source version differs.
prepare_data.py reconstructs the tables. verify_data.py is a separate
streaming implementation using only the Python standard library and the
original ZIP, and checks the produced tables without importing pandas logic.

No trade entries, fills, returns, success rates or profitability are calculated.
