Metadata-Version: 2.4
Name: ptcna
Version: 0.1.0
Summary: Prime Tensor Circled Neural Architecture — one architecture, four layers (neural/circle/seed/core). Consolidates pcna, pcta, pcsa.
Author-email: Erin Patrick Spencer <wayseer@interdependentway.org>
License-Expression: MPL-2.0
Project-URL: Homepage, https://github.com/The-Interdependency/ptcna
Project-URL: Repository, https://github.com/The-Interdependency/ptcna
Project-URL: Bug Tracker, https://github.com/The-Interdependency/ptcna/issues
Keywords: prime,tensor,neural-architecture,back-propagation,auditing,timing,fick,unit-circle,interdependency
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Science/Research
Classifier: Intended Audience :: Developers
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Topic :: Software Development :: Libraries
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: numpy>=1.21
Provides-Extra: dev
Requires-Dist: pytest>=8.0; extra == "dev"
Requires-Dist: build; extra == "dev"
Requires-Dist: twine; extra == "dev"
Dynamic: license-file

# PTCNA — Prime Tensor Circled Neural Architecture

One architecture, four layers. Not four repos — four **layers of one thing**.
PTCNA consolidates the former The-Interdependency repos `pcna`, `pcta`, and
`pcsa` into a single package and is the single upstream that feeds
[`interdependent-lib`](https://github.com/The-Interdependency/interdependent-lib).

## The four layers

Each layer's tensors divide into the next; every circle, seed, and core **is
itself a tensor**.

| Module | Layer | Divides… → … | Tensor kind | Back-propagation |
|---|---|---|---|---|
| `ptcna.neural` | neural | (base) neural tensors | **neural** | **yes — the only differentiable layer** |
| `ptcna.circle` | circle | neural tensors → circles | auditing / timing | no |
| `ptcna.seed` | seed | circles → seeds | auditing / timing | no |
| `ptcna.core` | core | seeds → cores | auditing / timing | no |

- **Back-propagation lives only in the neural layer.** Circle, seed, and core
  tensors are **auditing and timing tensors** — they do not differentiate.
- **fiqs** gate *when cores propagate internally*, per Fick's first law
  `J = −D ∇φ` (structure diffusing down its field gradient). Timing, not
  gradient descent. The fiq substrate lives in `ptcna.core.prime_core`.
- **PCEA** (Prime Circular Encryption Algorithm) is **not** a layer — it stays a
  separate, orthogonal repo (the guardian: "last state as key" at every layer).

## Provenance

| Layer | Migrated from | Was |
|---|---|---|
| neural | `The-Interdependency/pcna` (`core/`) | Prime Circular Neural Architecture |
| seed | `The-Interdependency/pcta` | Prime Circled Tensor Architecture (circles → seeds) |
| core | `The-Interdependency/pcsa` (`ptca/` + `prime_core/`) | Prime Tensor Core Architecture (was `PTCA`) |
| circle | *new* (extraction target — see `docs/architecture.md`) | previously unnamed |

## Install & test

```bash
pip install -e ".[dev]"      # neural layer needs numpy; seed/core are stdlib-only
pytest                       # testpaths = ptcna
```

## Status

Alpha (`0.1.0`). All four layers import; **146 tests pass** (seed/core/prime_core
stdlib-only + neural under numpy). Consolidation reconciliation is complete for
the neural/circle/seed split: the seed/circle audit was extracted into its
layers and the neural prime-ring tensor was renamed `PTCACore` → `RingCore`.
The core layer still intentionally exposes PTCA-named public objects such as
`PTCATensor` and `PTCAInstance`; those names now live in the correct layer.
History and any remaining notes live in `docs/architecture.md`.

License: MPL-2.0.
