feat: v0.4.0 package restructuring — split into 4 modules

Separate monolithic beta_spectrum package into:
  beta_spectrum   — pure theory (components, spectrum, CLI, visualize)
  exp_data        — experimental data handling (spectrum, calibration,
                    fitters, corrections)
  fitter          — analysis framework (model, fit_engine, extractor,
                    result)
  beta_spectrum.visualize — plotting (BetaSpectrumAnalyzer)

Phase 1 — exp_data module:
  ExpSpectrum dataclass with metadata, normalization, serialization
  EnergyCalibrator (linear/quadratic calibration)
  GaussianFitter + PeakFitter for calibration sources
  DeadTimeCorrection, PileUpCorrection, BackgroundSubtractor
  43 new tests, ruff + mypy clean

Phase 2 — fitter module:
  SpectrumModel wraps BetaSpectrum + DetectorResponse with convolution
  SpectrumFitter with χ² minimization, covariance, profile likelihood,
  confidence intervals, robust loss functions
  CWExtractor + GVAExtractor for C(W) and g_V/g_A extraction
  AnalysisResult + FitSummary for physics-specific results
  CurveFitter general-purpose callable model fitter
  34 new tests, ruff + mypy clean

Phase 3 — clean up beta_spectrum:
  Remove cw_extractor.py and fitter.py (moved to fitter/)
  Clean __init__.py to export only theory classes
  Add backward-compat methods to CWExtractor
  Fix BetaSpectrumAnalyzer._get_branch_spectrum returning BetaSpectrum
  objects instead of numpy arrays (plot fix)

Phase 4 — package structure:
  Auto-discovered by setuptools (4 packages)
  All imports updated across 26 source files
  Black formatting applied
  322 tests pass (262 original + 60 new)
  ruff + mypy clean on all source files

Bug fixes:
  - Fix multi-branch spectrum plotting: universal component keys
    (Fermi, Screening, etc.) don't have per-branch prefix; fixed
    fallback lookup.
  - Fix negative phase space beyond endpoint: BetaSpectrum returns
    negative values for W > W0; clamp branch spectra to zero.

tests/exp_data/test_fitters.py

@@ -0,0 +1,114 @@
+"""Tests for exp_data.fitters."""
+
+import numpy as np
+
+from exp_data.fitters import GaussianFitter, PeakFitter
+
+
+class TestGaussianFitter:
+    """Test Gaussian peak fitting."""
+
+    def _make_synthetic_peak(
+        self,
+        center=50.0,
+        sigma=2.0,
+        amplitude=100.0,
+        background=10.0,
+        n_points=500,
+        energy_range=(0, 100),
+    ):
+        """Create synthetic Gaussian peak with noise."""
+        energies = np.linspace(energy_range[0], energy_range[1], n_points)
+        counts = (
+            amplitude * np.exp(-0.5 * ((energies - center) / sigma) ** 2) + background
+        )
+        counts += np.random.RandomState(42).normal(
+            0, np.sqrt(np.maximum(counts, 1)), n_points
+        )
+        return energies, counts
+
+    def test_fit_returns_valid_result(self):
+        energies, counts = self._make_synthetic_peak()
+        fitter = GaussianFitter(energies, counts)
+        result = fitter.fit()
+
+        assert result.success
+        assert result.chi2 > 0
+        assert result.n_dof > 0
+        assert result.covariance.shape == (4, 4)
+
+    def test_fit_recovers_parameters(self):
+        center, sigma, amplitude, background = 50.0, 2.0, 100.0, 10.0
+        energies, counts = self._make_synthetic_peak(
+            center=center, sigma=sigma, amplitude=amplitude, background=background
+        )
+        fitter = GaussianFitter(energies, counts)
+        result = fitter.fit()
+
+        np.testing.assert_allclose(result.center_keV, center, rtol=0.05)
+        np.testing.assert_allclose(result.sigma_keV, sigma, rtol=0.1)
+        np.testing.assert_allclose(result.amplitude, amplitude, rtol=0.15)
+
+    def test_resolution_property(self):
+        energies, counts = self._make_synthetic_peak(sigma=2.0)
+        fitter = GaussianFitter(energies, counts)
+        result = fitter.fit()
+
+        expected_fwhm = 2.35482 * 2.0
+        np.testing.assert_allclose(result.resolution(), expected_fwhm, rtol=0.05)
+
+    def test_resolution_fraction(self):
+        energies, counts = self._make_synthetic_peak(center=50.0, sigma=2.0)
+        fitter = GaussianFitter(energies, counts)
+        result = fitter.fit()
+
+        expected_frac = (2.35482 * 2.0) / 50.0
+        np.testing.assert_allclose(
+            result.resolution_fraction(), expected_frac, rtol=0.05
+        )
+
+    def test_auto_guess(self):
+        energies, counts = self._make_synthetic_peak()
+        fitter = GaussianFitter(energies, counts)
+        result = fitter.fit()
+        assert result.success
+
+
+class TestPeakFitter:
+    """Test multi-peak detection and fitting."""
+
+    def _make_multi_peak(self):
+        """Create spectrum with two peaks."""
+        energies = np.linspace(0, 100, 1000)
+        counts = np.zeros_like(energies)
+        counts += 50 * np.exp(-0.5 * ((energies - 20) / 2) ** 2)
+        counts += 80 * np.exp(-0.5 * ((energies - 70) / 3) ** 2)
+        counts += 5  # constant background
+        counts += np.random.RandomState(42).normal(
+            0, np.sqrt(np.maximum(counts, 1)), len(energies)
+        )
+        return energies, counts
+
+    def test_find_peaks(self):
+        energies, counts = self._make_multi_peak()
+        fitter = PeakFitter(energies, counts)
+        peaks = fitter.find_peaks(min_height=10.0, min_distance_keV=10.0)
+
+        assert len(peaks) >= 1  # At least one peak should be found
+
+    def test_fit_peaks_with_known_energies(self):
+        energies, counts = self._make_multi_peak()
+        fitter = PeakFitter(energies, counts)
+        results = fitter.fit_peaks(known_energies=np.array([20.0, 70.0]))
+
+        assert len(results) >= 1
+        for r in results:
+            assert r.success
+            assert r.n_dof > 0
+
+    def test_fit_peaks_auto_detect(self):
+        energies, counts = self._make_multi_peak()
+        fitter = PeakFitter(energies, counts)
+        results = fitter.fit_peaks()
+
+        assert len(results) >= 1