Data — GCPI Research

Primary releases

CN research indices

Weekly series regenerated by scripts/build_cn_indices.py from ComputeNav prices_YYYY-MM-DD.csv panels. Bundle: api/v1/indices.json.

↓ indices.json ↓ cn_token_spot.csv ↓ cn_frontier.csv ↓ cn_open.csv ↓ cn_arb_spread.csv

File	Contents
indices.json	All live series + meta
cn_token_spot.csv	Floor / P25 / Median USD, wow, optional index_*
cn_frontier.csv	Geo-mean level USD, seats, wow, index_level
cn_open.csv	Geo-mean level USD, groups, wow, index_level
cn_arb_spread.csv	Median relative spread %, range USD, groups

Academic GCPI (v0.1)

Inaugural benchmark files

Separate from CN-Token Spot. Used by scripts/compute_gcpi.py for the standardised academic headline.

↓ prices_2026-04.csv ↓ constituents_2026-04-a.csv ↓ gcpi_timeseries.csv ↓ compute_gcpi.py ↓ build_cn_indices.py

Reproduce

Build commands

# From Web-GCPI Research repository root
python3 scripts/build_cn_indices.py
# → data/processed/cn_{token_spot,frontier,open,arb_spread}.csv
# → api/v1/indices.json
# → optional mirror to sibling ComputeNav api/v1/cn_indices.json

Requires the sibling Web-ComputeNav tree (or pass --prices-dir). Stdlib only — no NumPy / pandas.