# NTS Voynich Empirical Suite (Local)

This folder contains production-ready scripts to replicate and extend the empirical validation
of the Voynich/NTS model. They accept the EVA corpus at: `/mnt/data/voinich EVA.txt`.

## Scripts

- `nts_prs_segment.py` — PRS segmentation of tokens.
- `nts_productivity.py` — Morphological productivity P and P_s per root.
- `nts_zipf_entropy.py` — Zipf exponent and Shannon entropy (+ optional plot).
- `nts_cooc.py` — Co-occurrence and n-gram extraction.
- `nts_random_check.py` — Falsifiability via controlled randomization.
- `nts_sts_full.py` — STS index and transition matrix on the full corpus.
- `nts_utils.py` — Common helpers (loading, checksums, segmentation, stats).

## Example run

```
mkdir -p /mnt/data/nts_outputs

python /mnt/data/nts_suite/nts_prs_segment.py --eva "/mnt/data/voinich EVA.txt"   --out "/mnt/data/nts_outputs/prs_segments.csv"   --report "/mnt/data/nts_outputs/prs_report.csv"

python /mnt/data/nts_suite/nts_productivity.py   --prs "/mnt/data/nts_outputs/prs_segments.csv"   --out "/mnt/data/nts_outputs/productivity.csv"

python /mnt/data/nts_suite/nts_zipf_entropy.py --eva "/mnt/data/voinich EVA.txt"   --out "/mnt/data/nts_outputs/zipf_entropy.json"   --plot "/mnt/data/nts_outputs/zipf_plot.png"

python /mnt/data/nts_suite/nts_cooc.py --eva "/mnt/data/voinich EVA.txt" --n 2   --top 1000 --out "/mnt/data/nts_outputs/cooc_bigram.csv"

python /mnt/data/nts_suite/nts_random_check.py --eva "/mnt/data/voinich EVA.txt"   --out "/mnt/data/nts_outputs/random_check.json" --seed 42

python /mnt/data/nts_suite/nts_sts_full.py --prs "/mnt/data/nts_outputs/prs_segments.csv"   --out "/mnt/data/nts_outputs/sts_summary.json"
```

## Notes
- Prefix/suffix sets are initialized from your manuscript and can be expanded in `nts_utils.py`.
- State mapping (S1..S5) follows your documented logic and can be refined as needed.
- All outputs are CSV/JSON/PNG for maximum portability.



# Core scripts (extended)

These three are the primary, extended scripts referenced in *Capitolo 12*:

- `nts_empirical_check.py` – extended single/multi-window empirical check (H, MI, P_s, STS), CSV + JSON.
  ```bash
  python /mnt/data/nts_suite/nts_empirical_check.py \
    --eva "/mnt/data/voinich EVA.txt" \
    --window 500 --windows 10 --shift 500 \
    --out_json "/mnt/data/nts_outputs/empirical_check.json" \
    --out_csv "/mnt/data/nts_outputs/empirical_check.csv" \
    --top 20
  ```

- `nts_eva_pie.py` – extended EVA–PIE similarity (configurable threshold, optional PIE JSON), CSV + JSON.
  ```bash
  python /mnt/data/nts_suite/nts_eva_pie.py \
    --eva "/mnt/data/voinich EVA.txt" \
    --min_sim 0.5 --top 200 \
    --out_csv "/mnt/data/nts_outputs/eva_pie_similarity.csv" \
    --out_json "/mnt/data/nts_outputs/eva_pie_summary.json"
  ```

- `nts_stability_s1s5.py` – extended stability metrics across windows (S1..S5), CSV + JSON.
  ```bash
  python /mnt/data/nts_suite/nts_stability_s1s5.py \
    --eva "/mnt/data/voinich EVA.txt" \
    --window 50 --windows 8 --shift 50 \
    --out_csv "/mnt/data/nts_outputs/stability_windows.csv" \
    --out_json "/mnt/data/nts_outputs/stability_summary.json"
  ```

These integrate cleanly with the rest of the suite already present in `/mnt/data/nts_suite`.