diff --git a/lea/app/run.py b/lea/app/run.py
index 2997287..3b39ec1 100644
--- a/lea/app/run.py
+++ b/lea/app/run.py
@@ -197,7 +197,9 @@ def run(
     whitelist = (
         # If multiple select queries are provided, we want to run the union of all the views they
         # select. We use a set union to remove duplicates.
-        set.union(*(make_whitelist(query, dag) for query in select)) if select else set(dag.keys())
+        set.union(*(make_whitelist(query, dag) for query in select))
+        if select
+        else set(dag.keys())
     )
     console_log(f"{len(whitelist):,d} view(s) selected")
 
diff --git a/lea/views/dag.py b/lea/views/dag.py
index be3b199..74e9236 100644
--- a/lea/views/dag.py
+++ b/lea/views/dag.py
@@ -48,19 +48,46 @@ def _list_descendants(node):
 
         return list(_list_descendants(node))
 
-    def _build_nested_schema(self, deps):
+    @property
+    def _nested_schema(self):
+        """
+
+        >>> import pathlib
+        >>> import lea
+        >>> from pprint import pprint
+
+        >>> views_dir = pathlib.Path(__file__).parent.parent.parent / "examples" / "jaffle_shop" / "views"
+        >>> views = lea.views.load_views(views_dir, sqlglot_dialect="duckdb")
+        >>> views = [view for view in views if view.schema not in {"tests"}]
+        >>> dag = lea.views.DAGOfViews(views)
+
+        >>> pprint(dag._nested_schema)
+        {'analytics': {'finance': {'kpis': {}}, 'kpis': {}},
+         'core': {'customers': {}, 'orders': {}},
+         'staging': {'customers': {}, 'orders': {}, 'payments': {}}}
+
+        """
+
+        nodes = set(node for deps in self.dependencies.values() for node in deps) | set(
+            self.dependencies.keys()
+        )
+
         nested_schema = {}
 
-        for path in deps:
+        for key in nodes:
             current_level = nested_schema
-            for part in path:
+            for part in key:
                 if part not in current_level:
                     current_level[part] = {}
                 current_level = current_level[part]
 
         return nested_schema
 
-    def _to_mermaid_subgraphs(self, out, nested_schema: dict):
+    def _to_mermaid_views(self):
+        out = io.StringIO()
+        out.write('%%{init: {"flowchart": {"defaultRenderer": "elk"}} }%%\n')
+        out.write("flowchart TB\n")
+
         def output_subgraph(schema: str, values: dict, prefix: str = ""):
             out.write(f"\n{FOUR_SPACES}subgraph {schema}\n")
             for value in sorted(values.keys()):
@@ -73,26 +100,19 @@ def output_subgraph(schema: str, values: dict, prefix: str = ""):
                     out.write(f"{FOUR_SPACES*2}{full_path}({value})\n")
             out.write(f"{FOUR_SPACES}end\n\n")
 
+        # Print out the nodes, within each subgraph block
+        nested_schema = self._nested_schema
         for schema in sorted(nested_schema.keys()):
             values = nested_schema[schema]
             output_subgraph(schema, values)
 
-    def _to_mermaid_views(self):
-        out = io.StringIO()
-        out.write('%%{init: {"flowchart": {"defaultRenderer": "elk"}} }%%\n')
-        out.write("flowchart TB\n")
-        nodes = set(node for deps in self.dependencies.values() for node in deps) | set(
-            self.dependencies.keys()
-        )
-
-        nested_schema = self._build_nested_schema(nodes)
-        self._to_mermaid_subgraphs(out, nested_schema)
-
+        # Print out the edges
         for dst, srcs in sorted(self.dependencies.items()):
             dst = ".".join(dst)
             for src in sorted(srcs):
                 src = ".".join(src)
                 out.write(f"{FOUR_SPACES}{src} --> {dst}\n")
+
         return out.getvalue()
 
     def _to_mermaid_schemas(self):