Dash web: crosshair, zoom/pan X, scale log/lin, continuum extraction, background resume

- Tooltip entier (intersect:false) + ligne verticale crosshair sur tous les graphes
- Zoom molette/pinch sur l'axe X, pan souris, limites clamped 30-3000 keV
- Toggle échelle log/linéaire onglet Background
- Extraction continuum détecteur (isotope peaks subtracted + Gaussian smoothing)
- Reprise snapshot précédent au démarrage capture_background.py
- Suppression refs "Théorique" et "Bruit capteur" de l'interface
- Plugin chartjs-plugin-zoom + hammerjs via CDN
- Fix Chart constructor spread operator

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

train/vega_ml/synthetic_spectra/physics/spectrum_physics.py

@@ -324,13 +324,16 @@ def generate_realistic_continuum(
     Generate realistic CsI(Tl) background continuum shape.
 
     Calibrated against real Radiacode 103 background measurements.
-    Produces the characteristic asymmetric hump at ~110 keV and
-    Compton-like tail that simple exponentials miss.
+    Produces the characteristic asymmetric hump at ~110 keV with
+    housing absorption at low energy, Compton plateau, and proper
+    high-energy falloff.
 
     Shape components:
-    - Asymmetric hump centered at ~110 keV (sigma_left=55, sigma_right=50 keV)
-    - Compton continuum: 0.45*exp(-E/240) + 0.04*exp(-E/700)
-    - Noise floor at 0.8% of peak
+    - Asymmetric hump at ~110 keV (sigma_left=48, sigma_right=80 keV)
+    - Housing absorption below ~40 keV: 1 - exp(-E/30)
+    - Compton plateau around 200-260 keV from Pb-214/Bi-214 scatter
+    - Compton tail: 0.38*exp(-E/170) + 0.06*exp(-E/500)
+    - Noise floor at 0.3% of peak
 
     Args:
         energy_bins: Array of energy bin centers (keV)
@@ -342,24 +345,31 @@ def generate_realistic_continuum(
     """
     E = energy_bins
 
-    # Asymmetric hump at ~110 keV (low-energy scatter peak in CsI(Tl))
+    # Asymmetric hump at ~110 keV
+    # Left side sharper (sigma=48), right side broader with Compton shoulder (sigma=80)
     hump_center = 110.0
-    sigma_left = 55.0   # Broader on the low-energy side
-    sigma_right = 50.0  # Narrower on the high-energy side
     hump = np.where(
         E <= hump_center,
-        np.exp(-0.5 * ((E - hump_center) / sigma_left) ** 2),
-        np.exp(-0.5 * ((E - hump_center) / sigma_right) ** 2),
+        np.exp(-0.5 * ((E - hump_center) / 48.0) ** 2),
+        np.exp(-0.5 * ((E - hump_center) / 80.0) ** 2),
     )
 
+    # Housing absorption at very low energy (< ~40 keV)
+    absorption = 1.0 - np.exp(-E / 30.0)
+
     # Compton continuum tail
-    tail = 0.45 * np.exp(-E / 240.0) + 0.04 * np.exp(-E / 700.0)
+    tail = 0.38 * np.exp(-E / 170.0) + 0.06 * np.exp(-E / 500.0)
 
-    # Noise floor (low-level baseline)
-    noise_floor = 0.008
+    # Compton plateau around 200-260 keV (Pb-214/Bi-214 scatter)
+    compton_edge = np.maximum(0, 1.0 - ((E - 180.0) / 150.0) ** 2)
+    compton_edge[E > 330] = 0
+    compton_plateau = 0.12 * compton_edge
 
-    # Combine shape components
-    continuum = hump + tail + noise_floor
+    # Noise floor
+    noise_floor = 0.003
+
+    # Combine continuum with absorption
+    continuum = (hump + tail + compton_plateau) * absorption + noise_floor
 
     # Normalize to target total counts
     if continuum.sum() > 0 and total_counts > 0: