fix(explorer): fix scree plot data and add bar+line combo chart

- Use all individual MPs (not party aggregates) for L2-norm computation;
  party-aggregated vectors have near-zero values on some dims due to
  Procrustes alignment, producing spurious zeros
- Sort importances descending so scree plot is properly monotonic
- Relabel x-axis as 'Rang' since dim ordering after Procrustes alignment
  no longer matches original singular value order
- Add Scatter line trace connecting bar tops for elbow visibility

explorer.py

@@ -252,23 +252,27 @@ def load_party_axis_scores(db_path: str) -> Dict[str, List[float]]:
 
 @st.cache_data(show_spinner="Scree-plot laden…")
 def load_scree_data(db_path: str) -> List[float]:
-    """Return a list of component importances (L2-norm of party scores per dimension).
+    """Return component importances (L2-norm per SVD dimension), sorted descending.
 
-    Uses the same svd_vectors data as load_party_axis_scores but aggregates across
-    all components (0-indexed). Returns a list of length == vector dimensionality (50).
+    Uses ALL individual MP vectors (entity_type='mp', window='current_parliament'),
+    excluding party-aggregated rows. Since the stored vectors are U*s (scaled by
+    singular values), the L2-norm of all MP scores per dimension approximates the
+    singular value for that dimension. Sorting descending gives the proper scree shape.
+
+    Note: Procrustes alignment across sub-windows may scramble the original dimension
+    ordering, so we sort by magnitude rather than relying on dimension index order.
     """
     try:
         con = duckdb.connect(database=db_path, read_only=True)
-        party_list = sorted(CURRENT_PARLIAMENT_PARTIES)
-        placeholders = ", ".join("?" for _ in party_list)
         rows = con.execute(
-            f"SELECT vector FROM svd_vectors "
-            f"WHERE entity_type='mp' AND window_id='current_parliament' "
-            f"AND entity_id IN ({placeholders})",
-            party_list,
+            "SELECT entity_id, vector FROM svd_vectors "
+            "WHERE entity_type='mp' AND window_id='current_parliament'"
         ).fetchall()
+        # Individual MPs have "Lastname, F." format; party rows are short codes without commas
         vectors: List[List[float]] = []
-        for (raw_vec,) in rows:
+        for entity_id, raw_vec in rows:
+            if "," not in entity_id:
+                continue  # skip party-aggregated rows
             if isinstance(raw_vec, str):
                 vec = json.loads(raw_vec)
             elif isinstance(raw_vec, (bytes, bytearray)):
@@ -289,7 +293,7 @@ def load_scree_data(db_path: str) -> List[float]:
             col = [v[dim] for v in vectors if dim < len(v)]
             l2 = sum(x**2 for x in col) ** 0.5
             importances.append(l2)
-        return importances
+        return sorted(importances, reverse=True)
     except Exception:
         logger.exception("Failed to load scree data")
         return []
@@ -301,33 +305,47 @@ def load_scree_data(db_path: str) -> List[float]:
 
 
 def _render_scree_plot(importances: List[float], n_show: int = 15) -> None:
-    """Render a bar chart showing relative component importance (scree plot).
+    """Render a bar+line combo chart showing relative SVD component importance.
+
+    Bars show the L2-norm (singular value proxy) per rank; a line connects the tops
+    of the bars to make the 'elbow' in the scree curve easy to spot.
 
     Args:
-        importances: List of L2-norm scores per component (0-indexed).
+        importances: List of importance values sorted descending (from load_scree_data).
         n_show: How many components to display (default: first 15).
     """
     if not importances:
         return
     data = importances[:n_show]
-    components = list(range(1, len(data) + 1))
-    colours = [
-        PARTY_COLOURS.get("PVV", "#1565C0") if i == 0 else "#90CAF9"
-        for i in range(len(data))
-    ]
-    fig = go.Figure(
+    ranks = list(range(1, len(data) + 1))
+    bar_colour = "#90CAF9"
+    line_colour = "#1565C0"
+    fig = go.Figure()
+    fig.add_trace(
         go.Bar(
-            x=components,
+            x=ranks,
             y=data,
-            marker_color=colours,
-            hovertemplate="As %{x}<br>Gewicht: %{y:.2f}<extra></extra>",
+            marker_color=bar_colour,
+            hovertemplate="Rang %{x}<br>Gewicht: %{y:.2f}<extra></extra>",
+            showlegend=False,
+        )
+    )
+    fig.add_trace(
+        go.Scatter(
+            x=ranks,
+            y=data,
+            mode="lines+markers",
+            line={"color": line_colour, "width": 2},
+            marker={"size": 6, "color": line_colour},
+            hoverinfo="skip",
+            showlegend=False,
         )
     )
     fig.update_layout(
         height=220,
         margin={"l": 10, "r": 10, "t": 10, "b": 30},
         xaxis={
-            "title": "SVD-as",
+            "title": "Rang",
             "tickmode": "linear",
             "tick0": 1,
             "dtick": 1,
@@ -342,6 +360,7 @@ def _render_scree_plot(importances: List[float], n_show: int = 15) -> None:
         },
         plot_bgcolor="rgba(0,0,0,0)",
         paper_bgcolor="rgba(0,0,0,0)",
+        bargap=0.2,
     )
     st.plotly_chart(fig, use_container_width=True)