Add functionality to plot a solution (#21)

* Add function to plot solution

* Add ability to plot street outlines

Add polyline lib to project

loggibud/v1/plotting/plot_solution.py

@@ -0,0 +1,163 @@
+"""Plots solution routes"""
+from typing import List, Iterable, Optional
+
+import folium
+import numpy as np
+import polyline
+import requests
+
+from loggibud.v1.types import CVRPSolution, Point
+from loggibud.v1.distances import OSRMConfig
+
+
+# All available map colors
+MAP_COLORS = (
+    "black",
+    "blue",
+    "darkred",
+    "purple",
+    "red",
+    "orange",
+    "green",
+    "pink",
+    "darkblue",
+    "beige",
+    "gray",
+    "lightgreen",
+    "lightblue",
+    "lightgray",
+    "cadetblue",
+)
+
+
+def plot_cvrp_solution_routes(
+    solution: CVRPSolution,
+    route_indices_to_plot: Optional[List[int]] = None,
+    config: Optional[OSRMConfig] = None,
+) -> None:
+    """Plot solution routes in a map along the streets
+
+    Parameters
+    ----------
+    solution
+        A solution to any solver with the vehicles routes to plot
+
+    route_indices_to_plot
+        If specified, selects a smaller subset of routes to plot by their
+        indices. This can be useful to reduce the clutter in case of a
+        solution with too many vehicles
+
+    config
+        OSRM configuration
+    """
+    config = config or OSRMConfig()
+
+    # Initialize map centered at the mean of the origins
+    origins_mean = np.mean(
+        [
+            (vehicle.origin.lat, vehicle.origin.lng)
+            for vehicle in solution.vehicles
+        ],
+        axis=0,
+    )
+    m = folium.Map(
+        location=origins_mean,
+        zoom_start=12,
+        tiles="cartodbpositron",
+    )
+
+    num_vehicles = len(solution.vehicles)
+    route_indices_to_plot = route_indices_to_plot or range(num_vehicles)
+    vehicles_subset = [solution.vehicles[i] for i in route_indices_to_plot]
+
+    for i, vehicle in enumerate(vehicles_subset):
+        vehicle_color = MAP_COLORS[i % len(MAP_COLORS)]
+
+        # Plot origin
+        origin = (vehicle.origin.lat, vehicle.origin.lng)
+        folium.CircleMarker(origin, color="red", radius=3, weight=5).add_to(m)
+
+        # Plot street outlines
+        wiring = _route_wiring(vehicle.circuit, config)
+        folium.PolyLine(
+            wiring, color=vehicle_color, weight=1.0, popup=f"Vehicle {i}"
+        ).add_to(m)
+
+        # Plot the deliveries as regular points
+        for delivery in vehicle.deliveries:
+            folium.Circle(
+                location=(delivery.point.lat, delivery.point.lng),
+                radius=10,
+                fill=True,
+                color=vehicle_color,
+                popup=(
+                    f"Vehicle {i} ({delivery.point.lat}, {delivery.point.lng})"
+                ),
+            ).add_to(m)
+
+    return m
+
+
+def _route_wiring(points: Iterable[Point], config):
+    coords_uri = ";".join(f"{point.lng},{point.lat}" for point in points)
+
+    response = requests.get(
+        f"{config.host}/route/v1/driving/{coords_uri}?overview=simplified",
+        timeout=config.timeout_s,
+    )
+
+    data = response.json()
+    line = data["routes"][0]["geometry"]
+
+    return [(lat, lng) for lat, lng in polyline.decode(line)]
+
+
+def plot_cvrp_solution(
+    solution: CVRPSolution, route_indices_to_plot: Optional[List[int]] = None
+) -> None:
+    """Plot solution deliveries in a map
+    This is a simplified version showing only the edges between each delivery.
+    It does not require an OSRM server configuration.
+
+    Parameters
+    ----------
+    solution
+        A solution to any solver with the vehicles routes to plot
+
+    route_indices_to_plot
+        If specified, selects a smaller subset of routes to plot by their
+        indices. This can be useful to reduce the clutter in case of a
+        solution with too many vehicles
+    """
+    # Initialize map centered at the mean of the origins
+    origins_mean = np.mean(
+        [
+            (vehicle.origin.lat, vehicle.origin.lng)
+            for vehicle in solution.vehicles
+        ],
+        axis=0,
+    )
+    m = folium.Map(
+        location=origins_mean,
+        zoom_start=12,
+        tiles="cartodbpositron",
+    )
+
+    num_vehicles = len(solution.vehicles)
+    route_indices_to_plot = route_indices_to_plot or range(num_vehicles)
+    vehicles_subset = [solution.vehicles[i] for i in route_indices_to_plot]
+
+    for i, vehicle in enumerate(vehicles_subset):
+        origin = (vehicle.origin.lat, vehicle.origin.lng)
+        folium.CircleMarker(origin, color="red", radius=3, weight=5).add_to(m)
+
+        vehicle_color = MAP_COLORS[i % len(MAP_COLORS)]
+        vehicle_coords = [(point.lat, point.lng) for point in vehicle.circuit]
+        folium.Polygon(
+            vehicle_coords,
+            popup=f"Vehicle {i}",
+            color=vehicle_color,
+            weight=1,
+        ).add_to(m)
+
+    return m

pyproject.toml

@@ -17,6 +17,7 @@ Shapely = "^1.7.1"
 pandas = "^1.2.3"
 geopandas = "^0.9.0"
 folium = "^0.12.1"
+polyline = "^1.4.0"
 
 [tool.poetry.dev-dependencies]
 pytest = "^6.2.3"