add: minor refactoring on significance and smaller note block in initial workflow

Author: BeaMarton13

doc/source/community_detection_guide.rst

@@ -58,7 +58,7 @@ This documentation provides an overview of the notebooks available in the commun
               </a>
           </div>
 
-   .. grid-item-card::  Test Significance of Community
+   .. grid-item-card::  Significance of Community Structure
       :link: community_detection_guide/notebooks/test_significance_of_community
       :link-type: doc
       :class-card: sd-card-hover card-yellow-orange

doc/source/community_detection_guide/notebooks/functions.ipynb

@@ -26,35 +26,36 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "def load_graph_from_gml_file(graph_file: str, weight_attribute_name: str = \"weight\"):\n",
-    "    import os\n",
-    "    # Check if the graph file exists\n",
-    "    if not os.path.exists(graph_file):\n",
-    "        print(f\"Error: Graph file '{graph_file}' not found.\")\n",
-    "        print(f\"Please ensure '{graph_file}' is a valid path to your 'lesmis.gml' file.\")\n",
-    "        print(\"You can typically find this file by searching for 'lesmis.gml network dataset'.\")\n",
-    "        return # Exit the function if file is not found\n",
-    "\n",
-    "    try:\n",
-    "        # Load network from GML file\n",
-    "        # igraph.Graph.Read_GML will automatically load edge attributes like 'value'\n",
-    "        # if they are present in the GML file.\n",
-    "        graph = ig.Graph.Read_GML(graph_file)\n",
+    "# def load_graph_from_gml_file(graph_file: str, weight_attribute_name: str = \"weight\"):\n",
+    "#     import os\n",
+    "#     # Check if the graph file exists\n",
+    "#     if not os.path.exists(graph_file):\n",
+    "#         print(f\"Error: Graph file '{graph_file}' not found.\")\n",
+    "#         print(f\"Please ensure '{graph_file}' is a valid path to your 'lesmis.gml' file.\")\n",
+    "#         print(\"You can typically find this file by searching for 'lesmis.gml network dataset'.\")\n",
+    "#         return # Exit the function if file is not found\n",
+    "\n",
+    "#     try:\n",
+    "#         # Load network from GML file\n",
+    "#         # igraph.Graph.Read_GML will automatically load edge attributes like 'value'\n",
+    "#         # if they are present in the GML file.\n",
+    "#         graph = ig.Graph.Read_GML(graph_file)\n",
     "        \n",
-    "        # Check if the graph has the correct weight attribute name\n",
-    "        if weight_attribute_name not in graph.edge_attributes():\n",
-    "            print(f\"Warning: Graph '{graph_file}' does not have a '{weight_attribute_name}' attribute. \"\n",
-    "                  \"Community detection will proceed without explicit weights, or if the algorithm \"\n",
-    "                  \"expects them, it might use default uniform weights.\")\n",
-    "            # If no 'value' attribute, assign a default uniform weight for visualization purposes\n",
-    "            graph.es[weight_attribute_name] = 1 \n",
+    "#         # Check if the graph has the correct weight attribute name\n",
+    "#         if weight_attribute_name not in graph.edge_attributes():\n",
+    "#             print(f\"Warning: Graph '{graph_file}' does not have a '{weight_attribute_name}' attribute. \"\n",
+    "#                   \"Community detection will proceed without explicit weights, or if the algorithm \"\n",
+    "#                   \"expects them, it might use default uniform weights.\")\n",
+    "#             # If no 'value' attribute, assign a default uniform weight for visualization purposes\n",
+    "#             graph.es[weight_attribute_name] = 1 \n",
     "\n",
-    "        return graph\n",
+    "#         return graph\n",
     "\n",
     "\n",
-    "    except Exception as e:\n",
-    "        print(f\"An error occurred while loading or processing the graph: {e}\")\n",
-    "        return\n",
+    "#     except Exception as e:\n",
+    "#         print(f\"An error occurred while loading or processing the graph: {e}\")\n",
+    "#         return\n",
+    "\n",
     "\n",
     "def community_detection(graph: ig.Graph, community_detection_method: str = \"multilevel\", weight_attribute_name: str = \"weight\", \n",
     "                        params: dict = None):\n",
@@ -67,9 +68,7 @@
     "            params[\"weights\"] = weight_attribute_name if weight_attribute_name in graph.edge_attributes() else None\n",
     "            return graph.community_leiden(**params)\n",
     "    elif community_detection_method == \"fastgreedy\":\n",
-    "        return graph.community_fastgreedy(weights=weight_attribute_name if weight_attribute_name in graph.edge_attributes() else None).as_clustering()\n",
-    "\n",
-    "\n"
+    "        return graph.community_fastgreedy(weights=weight_attribute_name if weight_attribute_name in graph.edge_attributes() else None).as_clustering()"
    ]
   },
   {
@@ -80,18 +79,21 @@
     "## Functions useful to test community structure"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "6309417d-2705-4454-9e68-84cf34883c1b",
+   "metadata": {},
+   "source": [
+    "### Method 1: Testing network structure with modularity"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": 3,
    "id": "1ab35a72-4bcd-4391-8f28-85b39c6a29ef",
    "metadata": {},
    "outputs": [],
    "source": [
-    "\n",
-    "def get_modularity_on_clustering(graph: ig.Graph, community_detection_method: str = \"multilevel\", params: dict = None):\n",
-    "    partition = community_detection(graph, community_detection_method, weight_attribute_name=None, params=params)\n",
-    "    return partition.modularity\n",
-    "\n",
     "def rewire(graph: ig.Graph, community_detection_method: str = \"multilevel\", params: dict = None):\n",
     "    num_randomizations = 500  # Number of randomized networks to generate\n",
     "    modularity_random_networks = []\n",
@@ -101,13 +103,23 @@
     "    for i in range(num_randomizations):\n",
     "        # G.rewire() modifies the graph in-place, so we must work on a copy.\n",
     "        graph_random = graph.copy()\n",
-    "    \n",
+    "\n",
     "        graph_random.rewire(n=num_swaps_for_randomization)\n",
-    "    \n",
-    "        modularity_random_networks.append(get_modularity_on_clustering(graph_random))\n",
+    "\n",
+    "        partition = community_detection(graph, community_detection_method, weight_attribute_name=None, params=params)\n",
+    "        modularity_random_networks.append(partition.modularity)\n",
     "\n",
     "    return modularity_random_networks\n",
     "\n",
+    "\n",
+    "def test_community_structure(graph: ig.Graph, graph_name: str = \"Karate Club Network\", community_detection_method: str = \"multilevel\", \n",
+    "                             params: dict = None):\n",
+    "    partition = community_detection(graph, community_detection_method, weight_attribute_name=None, params=params)\n",
+    "    modularity_orig = partition.modularity\n",
+    "    modularity_random_networks = rewire(graph, community_detection_method, params)\n",
+    "    plot_histogram(modularity_orig, modularity_random_networks, graph_name)\n",
+    "    \n",
+    "\n",
     "def plot_histogram(modularity_original: float, modularity_random_networks: list[float], graph_name: str=\"Karate Club Network\"):\n",
     "    import matplotlib.pyplot as plt\n",
     "    \n",
@@ -128,14 +140,15 @@
     "    plt.legend()\n",
     "    plt.grid(axis='y', alpha=0.75)\n",
     "    plt.tight_layout()\n",
-    "    plt.show()\n",
-    "\n",
-    "\n",
-    "def test_community_structure(graph: ig.Graph, graph_name: str = \"Karate Club Network\", community_detection_method: str = \"multilevel\", \n",
-    "                             params: dict = None):\n",
-    "    modularity_orig = get_modularity_on_clustering(graph, community_detection_method, params)\n",
-    "    modularity_random_networks = rewire(graph, community_detection_method, params)\n",
-    "    plot_histogram(modularity_orig, modularity_random_networks, graph_name)\n"
+    "    plt.show()\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "7eba4c89-d320-4178-bf00-77966010d148",
+   "metadata": {},
+   "source": [
+    "### Method 2: Testing network structure with NMI values"
    ]
   },
   {
@@ -313,23 +326,22 @@
     "    return pairwise_nmi_values\n"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "45c898fe-e73d-4b0d-aed2-6574df05132d",
+   "metadata": {},
+   "source": [
+    "### Testing Significance of Community Structure on a Grid Graph"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": 5,
    "id": "0e018281-2fe7-4c7c-b608-c2e97e315e34",
    "metadata": {},
    "outputs": [],
    "source": [
-    "def create_grid_graph(rows, cols, circular=False):\n",
-    "    \"\"\"\n",
-    "    Generates a 2D grid graph (lattice) without plotting.\n",
-    "    Returns the graph and its column count for later layout.\n",
-    "    \"\"\"\n",
-    "    G = ig.Graph.Lattice(dim=[rows, cols], nei=1, circular=circular)\n",
-    "    return G\n",
-    "\n",
-    "# --- Final version of cluster_and_plot_leiden with grid_cols ---\n",
-    "def cluster_and_plot_leiden_on_grid(graph, grid_cols, title=\"Graph with Leiden Communities\", plot_size=(8, 8)):\n",
+    "def plot_leiden_on_grid(graph, grid_cols, communities, title=\"Graph with Leiden Communities\", plot_size=(8, 8)):\n",
     "    \"\"\"\n",
     "    Clusters a graph using the Leiden algorithm and plots the result\n",
     "    with vertices colored by their community, specifically for grid layouts.\n",
@@ -344,11 +356,6 @@
     "        igraph.clustering.VertexClustering: The community detection result.\n",
     "    \"\"\"\n",
     "    import matplotlib.pyplot as plt\n",
-    "    \n",
-    "    print(f\"Clustering graph with {graph.vcount()} vertices and {graph.ecount()} edges using Leiden algorithm...\")\n",
-    "\n",
-    "    resolution = 0.15\n",
-    "    communities = graph.community_leiden(objective_function=\"modularity\", resolution=resolution)\n",
     "\n",
     "    # Assign colors based on community membership\n",
     "    palette = ig.GradientPalette(\"red\", \"blue\", n=len(communities)) \n",
@@ -376,9 +383,7 @@
     "    )\n",
     "    ax.set_title(title)\n",
     "    ax.axis('off')\n",
-    "    plt.show()\n",
-    "\n",
-    "    return communities, resolution"
+    "    plt.show()"
    ]
   },
   {