Four standalone Python scripts — fully reproducible, zero proprietary data
All scripts use the Sefaria.org API (public domain). No proprietary databases, no manual annotations, no hidden steps.
No external packages required. Standard library only. Any researcher can run these scripts immediately.
Every statistical finding in the book can be reproduced from these four scripts. Numbers match exactly.
The core algorithm. Classifies all 22 Hebrew letters into four groups (Foundation, AMTN, YHW, BKL). Computes Foundation% for any word, verse, or passage. Extracts mandatory roots. Detects trapped YHW letters. Runs Z-score shuffle tests.
Key results: Z = 152.16 · 99.87% dominance · 87.8% meaning prediction
python3 torah_root_analyzer.py שדי פרה אפר נחש
python3 torah_root_analyzer.py --zscore
python3 torah_root_analyzer.py --demo
Cross-validated meaning prediction using the root + YHW model. Tests whether Foundation root identity plus YHW letter context can predict semantic group membership. 5-fold cross-validation on 98,122 word pairs.
Key results: 87.8% accuracy (root+YHW) · 92.1% with nikud · 85.4% root-only baseline
python3 torah_meaning_predictor.py --cv
python3 torah_meaning_predictor.py --nikud
Variance scaling analysis, autocorrelation function, power spectrum computation, and terrain visualization. Tests whether Foundation% exhibits power-law correlations characteristic of physical systems at criticality.
Key results: ξ ≈ 1,104 verses · Spectral peaks at 254, 450, 1,169, 2,923 · α = -0.47
python3 torah_letter_flow.py --full
python3 torah_letter_flow.py --spectrum
Extracts genealogical trees from raw Torah text using nine structural parsing rules. Identifies parent-child relationships, tribal affiliations, and lineage chains across Genesis through Deuteronomy.
python3 torah_genealogy_parser.py --genesis
python3 torah_genealogy_parser.py --full
CC BY 4.0 · 1,393 total lines · 4 files