docs: update README with CLI, nuclear data integration, and additional modules

This commit is contained in:
2026-05-02 22:23:37 +03:00
parent 1269244b93
commit d8c521f6f8
+27 -121
View File
@@ -48,6 +48,10 @@ beta_shape_pnpi/
│ ├── constants.py # Physical constants (natural units, m_e = 1)
│ ├── utils.py # Helpers: T<->W conversion, momentum, nuclear radius
│ ├── spectrum.py # BetaSpectrum + BetaSpectrumAnalyzer classes
│ ├── nuclear_data.py # paceENSDF integration + JSON input support
│ ├── cli.py # Command-line interface (bs_pnpi)
│ ├── fitter.py # χ² curve fitting for C(W) extraction
│ ├── cw_extractor.py # C(W) shape factor extraction + gV/gA analysis
│ └── components/
│ │ ├── phase_space.py ✓ Phase space shape (p·W·(W₀−W)²)
│ │ ├── fermi.py ✓ Coulomb correction (loggamma for stability)
@@ -55,17 +59,23 @@ beta_shape_pnpi/
│ │ ├── screening.py ✓ Atomic electron screening (ratio method)
│ │ ├── exchange.py ✓ Hayen-2018 empirical fit coefficients
│ │ ├── radiative.py ✓ Outer radiative corrections, soft-photon resummation
│ │ ├── detector_response.py ✓ Analytical detector response models
│ │ └── recoil.py ✗ Not yet implemented
├── tests/
│ ├── test_radiative.py # Radiative correction tests (24 tests)
│ ├── test_exchange.py # Exchange correction tests
│ ├── test_fermi.py # Fermi function tests
│ ├── test_finite_size.py # Finite size correction tests
│ ├── test_phase_space.py # Phase space tests
│ ├── test_screening.py # Screening correction tests
│ └── test_spectrum.py # Integration tests
│ ├── test_radiative.py # Radiative correction tests (24 tests)
│ ├── test_exchange.py # Exchange correction tests
│ ├── test_fermi.py # Fermi function tests
│ ├── test_finite_size.py # Finite size correction tests
│ ├── test_phase_space.py # Phase space tests
│ ├── test_screening.py # Screening correction tests
│ ├── test_spectrum.py # Integration tests
│ ├── test_detector_response.py# Detector response tests
│ ├── test_fitter.py # Curve fitter tests
│ ├── test_cw_extractor.py # C(W) extraction tests
│ └── test_nuclear_data.py # paceENSDF + JSON input tests (23 tests)
└── data/
└── exchange_coeff.csv # Tabulated coefficients for X(Z,W), Z=2..120
├── exchange_coeff.csv # Tabulated coefficients for X(Z,W), Z=2..120
└── custom_input_example.json # Sample JSON input file
```
______________________________________________________________________
@@ -133,122 +143,18 @@ ______________________________________________________________________
| Phase Space | `phase_space.py` | ✓ | Baseline $p W (W_0 - W)^2$ with transition-type-dependent forbidden factors and optional neutrino mass support |
| Fermi Function | `fermi.py` | ✓ | Relativistic Coulomb correction via `scipy.special.loggamma` |
| Finite Size L0 | `finite_size.py` | ✓ | Low-Z expansion from Hayen et al. with prefactor |
| Charge Distribution U | `finite_size.py` | ✓ | Second-order $(1/5)(\\alpha Z W R)^2$ correction |
| Charge Distribution U | `finite_size.py` | ✓ | Second-order $(1/5)(\alpha Z W R)^2$ correction |
| Screening | `screening.py` | ✓ | Ratio method with logistic switching, energy shift $V_0$ |
| Exchange | `exchange.py` | ✓ | Hayen-2018 Table X empirical fit (CSV data for Z=2..120) |
| Radiative | `radiative.py` | ✓ | Outer radiative corrections (Hayen Eq. 47–53), soft-photon resummation at endpoint |
| Detector Response | `detector_response.py` | ✓ | Analytical models: Gaussian, Gaussian+tail, Tikhonov, tabulated |
| Recoil | `recoil.py` | ✗ | Stub only, not yet implemented |
______________________________________________________________________
## Additional Modules
## Configuration
Use `SpectrumConfig` to selectively enable/disable each correction:
```python
config = SpectrumConfig(
Z_parent=20, Z_daughter=21, A_number=40, endpoint_MeV=5.0,
transition_type="F1", # first forbidden non-unique
use_fermi=True,
use_screening=False, # disable atomic screening
use_exchange=True,
use_radiative=True,
)
```
______________________________________________________________________
## Transition Types
The `transition_type` parameter in `SpectrumConfig` determines the **forbidden factor** applied to the phase space calculation. This encodes the angular momentum and parity selection rules for the nuclear transition:
| Type | Name | Forbidden Order | $\\Delta J^\\pi$ | Forbidden Factor |
| ----- | ----------------------------- | --------------- | ------------------------------------- | --------------------------------------------------------- |
| `A` | Allowed | 0 | $0^+, 1^+$ | $1$ |
| `F1` | First forbidden (non-unique) | 1 | $0^-, 1^-, 2^\\pm$ | $p\_\\nu^2 + p_e^2$ |
| `F1U` | First forbidden (unique) | 1 | $2^-$ | $p\_\\nu^2 + p_e^2$ |
| `F2` | Second forbidden (non-unique) | 2 | $1^\\pm, 2^+, 3^\\pm$ | $p\_\\nu^4 + \\frac{3}{10}p\_\\nu^2 p_e^2 + p_e^4$ |
| `F2U` | Second forbidden (unique) | 2 | $3^-$ | $p\_\\nu^4 + \\frac{3}{10}p\_\\nu^2 p_e^2 + p_e^4$ |
| `F3` | Third forbidden (non-unique) | 3 | $0^\\pm, 1^\\pm, 2^\\pm, 3^+, 4^\\pm$ | $p\_\\nu^6 + 7p\_\\nu^4 p_e^2 + 7p\_\\nu^2 p_e^4 + p_e^6$ |
| `F3U` | Third forbidden (unique) | 3 | $4^-$ | $p\_\\nu^6 + 7p\_\\nu^4 p_e^2 + 7p\_\\nu^2 p_e^4 + p_e^6$ |
| `F4` | Fourth forbidden (unique) | 4 | $5^-$ | $p\_\\nu^6 + 7p\_\\nu^4 p_e^2 + 7p\_\\nu^2 p_e^4 + p_e^6$ |
The forbidden factor multiplies the baseline phase space $p_e W_e p\_\\nu W\_\\nu$, modifying the spectral shape at higher energies. The even/odd suffix (`U`) distinguishes **unique** transitions (single contributing nuclear matrix element) from **non-unique** ones (multiple contributing matrix elements).
> [!note] Shape factor
> For a complete treatment of forbidden transitions, the nuclear **shape factor** $C(Z,W)$ from `[[10-nuclear-structure]]` should also be included. This is not yet implemented in the calculator.
## Dependencies
**Runtime:** numpy, scipy (for `loggamma`, `spence`), matplotlib (plotting), pandas (CSV export)\
**Dev:** pytest, black, ruff, mypy (strict mode configured in `pyproject.toml`)
______________________________________________________________________
## Design Principles
### 1. Modularity
Each physical effect is isolated in its own component under `beta_spectrum/components/`.
### 2. Numerical Stability
Care near $W \\to 1$ (threshold) and $W \\to W_0$ (endpoint):
- `np.clip` / epsilon floors for regularization
- Safe logarithms via `scipy.special.loggamma` and dilogarithm
- Masking invalid regions
### 3. Vectorization
All components operate on NumPy arrays — no Python loops in hot paths.
### 4. Extensibility
New corrections are added by subclassing:
```python
class NewCorrection(SpectrumComponent):
def __call__(self, W):
return ...
```
______________________________________________________________________
## Development Status
**Version:** 0.2.0\
**Status:** Active development — core physics engine is functional with comprehensive test coverage.
### Remaining Work
- Implement nuclear recoil correction in `recoil.py` (see \[[07-recoil-effects]\])
- Declare pandas as a runtime dependency in `pyproject.toml`
- Satisfy mypy strict type-checking requirements
- CLI entry point and multiple decay branch support
- Implement nuclear shape factors C(Z,W) for forbidden transitions
- Implement neutrino radiative corrections (Sirlin 2011)
______________________________________________________________________
## Knowledge Base
The [`docs/`](docs/) directory contains the complete scientific reference for beta spectrum shape corrections, organized as an Obsidian vault. It is structured around Hayen et al., *Rev. Mod. Phys.* **90**, 015008 (2017) and maps directly to calculator components:
| # | Note | Calculator Module | Status |
| --- | ------------------------------ | --------------------------- | --------------------- |
| 1 | \[[01-fermi-function]\] | `components/fermi.py` | ✓ implemented |
| 2 | \[[03-finite-size]\] | `components/finite_size.py` | ✓ implemented |
| 3 | \[[04-screening-correction]\] | `components/screening.py` | ✓ implemented |
| 4 | \[[05-exchange-correction]\] | `components/exchange.py` | ✓ implemented |
| 5 | \[[06-radiative-corrections]\] | `components/radiative.py` | ✓ implemented |
| 6 | \[[07-recoil-effects]\] | `components/recoil.py` | ✗ not yet implemented |
Notes prefixed with `10-`, `12-`, `13-` cover nuclear structure, atomic overlap, and chemical effects — important for future extensions but outside the current calculator scope.
______________________________________________________________________
## Notes
- Physical constants are defined in natural units ($m_e = 1$); energies can be converted from MeV via utilities in `utils.py`
- The framework assumes familiarity with beta decay theory
| Module | Description |
| --------------------- | -------------------------------------------------------------------------------------------------------------- |
| `nuclear_data.py` | paceENSDF integration + JSON input for automated `SpectrumConfig` generation |
| `cli.py` | Command-line interface (`bs_pnpi`) for quick spectrum calculations |
| `fitter.py` | χ² curve fitting for C(W) shape factor extraction from experimental data |
| `cw_extractor.py` | C(W) extraction pipeline with gV/gA coupling constant analysis |