| File | Description |
|---|---|
sample/cc10plus_roots_sample_1000.txt |
First 1000 roots from the full dataset (for testing scripts) |
gap_closest.csv |
500 tightest root pairs with hex values |
gap_comprehensive.csv |
Per-(CC, bits) gap statistics |
gap_heatmap.csv |
Gap distribution bucketed by 2^X |
gap_pyramid.csv |
Gap distribution per CC level |
gap_statistics.csv |
Summary statistics per bucket |
cc_twins.csv |
Top CC-twin pairs (same-CC and cross-CC) |
kill_positions.csv |
Chain-breaker positions by small prime, per (CC, bits) bucket |
immunization_heatmap.csv |
Residue immunity and safe-residue rates per (CC, bits) bucket |
The full dataset (cc10plus_roots_snapshot_2026-03-12.txt.gz, ~30 MB uncompressed) is available as a GitHub Release asset. It contains 1,773,994 roots in the format:
CC10 0xHEXVALUE DIGITS
CC11 0xHEXVALUE DIGITS
...
Dataset summary:
| CC level | Count |
|---|---|
| CC10 | 749,024 |
| CC11 | 132,580 |
| CC12 | 39,138 |
| CC13 | 7,226 |
| CC14 | 1,323 |
| CC15 | 238 |
| CC16 | 44 |
| CC17 | 1 |
Primary search band: 89-91 bits. Additional data from 59-153 bit ranges.
kill_positions.csv — Where chains break, by small prime.
| Column | Meaning |
|---|---|
cc |
Official chain length of the root |
bits |
Bit size of the root |
prime |
Small prime that divides the composite link |
position |
Chain position of the break (1 = first link after the root, 2 = second, etc.) |
count |
Number of roots in this bucket killed at this position by this prime |
pct_of_bucket |
Percentage of the (cc, bits) bucket |
immunization_heatmap.csv — Residue immunity summary per (cc, bits) bucket.
imm_<p> / imm_<p>_pct: count and percentage of roots immune to prime p (i.e., p cannot kill any chain position).safe_<p> / safe_<p>_pct: count and percentage of roots where the residue mod p is “safe” — the root avoids the dangerous residue classes for that prime.safe_all / safe_all_pct: roots that satisfy all tracked safe-residue conditions simultaneously.Both CSVs are derived analysis outputs generated from the released dataset snapshot, not hand-maintained files.
# Gap analysis (requires dataset file)
python3 analysis/scripts/cc_gap_analysis.py cc10plus_roots_snapshot_2026-03-12.txt --prefix gap --min-cc 10
# Ghost chain scan (~12 min on 8 cores)
python3 analysis/scripts/ghost_chains.py cc10plus_roots_snapshot_2026-03-12.txt
# Immunization analysis
python3 analysis/scripts/analyze_cc_immunization.py cc10plus_roots_snapshot_2026-03-12.txt