added calculate_network_statistics to functions.py and implemented it in linkbikenet.py

LinkBikeNet_MVP.ipynb

@@ -141,7 +141,7 @@
     "closest_pairs = []\n",
     "for i in range(to_iterate):\n",
     "    wcc = [H.subgraph(c).copy() for c in sorted(nx.connected_components(H), key=lambda c: sum([l[-1] for l in H.subgraph(c).copy().edges.data('length')]), reverse=True)]\n",
-    "    pair = pair_between_closest_components(wcc)\n",
+    "    pair = pair_between_largest_components(wcc)\n",
     "    closest_pairs.append(pair)\n",
     "    H.add_edge(pair[0], pair[1], length=0)"
    ],
@@ -201,6 +201,48 @@
    "outputs": [],
    "execution_count": null
   },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "source": "len(gdf)",
+   "id": "88be0fda819f9012",
+   "outputs": [],
+   "execution_count": null
+  },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "source": [
+    "# reset Graph\n",
+    "H = G.edge_subgraph(edges).copy()\n",
+    "\n",
+    "# calculating connectivity metrics\n",
+    "print(\"Calculating connectivity metrics...\")\n",
+    "network_lengths = []\n",
+    "lcc_lengths = []\n",
+    "edge_lengths = gdf['geometry'].length\n",
+    "\n",
+    "for i in range(len(gdf)):\n",
+    "    H.add_edge(closest_pairs[i][0], closest_pairs[i][1], length=edge_lengths[i])\n",
+    "    total, largest = calculate_network_statistics(H)\n",
+    "    network_lengths.append(total)\n",
+    "    lcc_lengths.append(largest)"
+   ],
+   "id": "fb40177941e881f0",
+   "outputs": [],
+   "execution_count": null
+  },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "source": [
+    "gdf['network_length'] = network_lengths\n",
+    "gdf['lcc_length'] = lcc_lengths"
+   ],
+   "id": "57d9c49a74f60f01",
+   "outputs": [],
+   "execution_count": null
+  },
   {
    "metadata": {},
    "cell_type": "code",

linkbikenet/functions.py

@@ -2,6 +2,7 @@
 from . import settings
 import re
 import osmnx as ox
+import networkx as nx
 import geopandas as gpd
 import numpy as np
 from scipy.spatial import cKDTree
@@ -358,4 +359,38 @@ def slugify(s):
         "aeiouaeiouaeiouaeiouaeiouaeiouaaaccilnnoorssssuyyzz"
     )
     s = s.translate(tab)
-    return s
+    return s
+
+def calculate_network_statistics(H):
+    """Return total network length and largest component length.
+
+    Parameters
+    ----------
+    H: networkx.Graph
+        undirected simple graph representing the street network with weighted edges
+
+    Returns
+    -------
+    total_length: float
+        total length of the network
+    largest_length: float
+        length of the largest connected component
+    """
+
+    total_length = sum(
+        data["length"]
+        for _, _, data in H.edges(data=True)
+    )
+    components = nx.connected_components(H)
+
+    largest_component = max(components, key=lambda c: sum(
+        H[u][v]["length"]
+        for u, v in H.subgraph(c).edges()
+    ))
+
+    largest_length = sum(
+        data["length"]
+        for _, _, data in H.subgraph(largest_component).edges(data=True)
+    )
+
+    return total_length, largest_length

linkbikenet/linkbikenet.py

@@ -157,6 +157,24 @@ def linkbikenet(
 
     gdf['ordering'] = gdf.index
 
+    # reset Graph
+    H = G.edge_subgraph(edges).copy()
+
+    # calculating connectivity metrics
+    print("Calculating connectivity metrics...")
+    network_lengths = []
+    lcc_lengths = []
+    edge_lengths = gdf['geometry'].length
+
+    for i in range(len(gdf)):
+        H.add_edge(closest_pairs[i][0], closest_pairs[i][1], length=edge_lengths[i])
+        total, largest = calculate_network_statistics(H)
+        network_lengths.append(total)
+        lcc_lengths.append(largest)
+
+    gdf['network_length'] = network_lengths
+    gdf['lcc_length'] = lcc_lengths
+
     edges_pbi_gdf = edges_gdf[edges_gdf["pbi"] == 1]
 
     # Back to unprojected (potentially). No more calculations after here.