matsim-proto/matsim_visualizer.py

import gzip

from lxml import etree
import networkx as nx
import matplotlib.pyplot as plt
import matplotlib.cm as cm
import matplotlib.colors as colors
from matplotlib.animation import FuncAnimation
from collections import defaultdict


class MatsimVisualizer:
  def __init__(self, input_file_path, output_file_path, viewport, show_facilities=False, show_nearby_nodes=False):
    self._nodes = {}
    self._links = []
    self._pos = None
    self.norm = None

    self._max_x = None
    self._min_x = None
    self._max_y = None
    self._min_y = None
    self._viewport = viewport

    self._output_file_path = output_file_path
    self._network_file_path = f"{input_file_path}/output_network.xml.gz"
    self._events_file_path = f"{input_file_path}/output_events.xml.gz"
    self._facilities_file_path = f"{input_file_path}/output_facilities.xml.gz"

    self._G = nx.Graph()
    self._traffic_per_tick = defaultdict(lambda: defaultdict(int))
    self._close_to_facility_nodes = set()

    self._tick = 0
    self._max_traffic = 0
    self._max_traffic_link = None
    self._max_traffic_tick = None

    self._show_facilities = show_facilities
    self._show_nearby_nodes = show_nearby_nodes

  def _load_data(self):
    # ====== LOAD FACILITIES ====== #
    if self._show_facilities:
      with gzip.open(self._facilities_file_path, 'rb') as file:
        facilities_doc = etree.parse(file)

      for node in facilities_doc.xpath('/facilities/facility'):
        x = float(node.get('x'))
        y = float(node.get('y'))
        if self._viewport[0] < x < self._viewport[2] and self._viewport[1] > y > self._viewport[3]:
          self._nodes[f"facility_{node.get('id')}"] = (x, y)

    # ====== LOAD NETWORK ====== #
    with gzip.open(self._network_file_path, 'rb') as file:
      network_doc = etree.parse(file)

    for node in network_doc.xpath('/network/nodes/node'):
      x = float(node.get('x'))
      y = float(node.get('y'))

      if self._max_x is None or x > self._max_x:
        self._max_x = x

      if self._min_x is None or x < self._min_x:
        self._min_x = x

      if self._max_y is None or y > self._max_y:
        self._max_y = y

      if self._min_y is None or y < self._min_y:
        self._min_y = y

      if self._viewport[0] < x < self._viewport[2] and self._viewport[1] > y > self._viewport[3]:
        self._nodes[node.get('id')] = (x, y)

    for link in network_doc.xpath('/network/links/link'):
      start_node = link.get('from')
      end_node = link.get('to')
      id = link.get('id')
      if start_node in self._nodes and end_node in self._nodes:
        self._links.append({
          'id': id,
          'from': start_node,
          'to': end_node,
          'vehicles': 0,
        })

    cumulative_traffic = defaultdict(int)

    # ====== LOAD EVENTS ====== #
    with gzip.open(self._events_file_path, 'rb') as file:
      events_doc = etree.parse(file)

    self._tick = 0
    for event in events_doc.xpath('/events/event'):
      link_id = event.get('link')
      event_type = event.get('type')
      time = float(event.get('time'))
      ticked = False

      if link_id is not None and event_type is not None and time is not None:
        if event_type in ['entered link', 'vehicle enters traffic']:
          self._traffic_per_tick[self._tick][link_id] += 1
          ticked = True

          cumulative_traffic[link_id] += 1
          # We need to find the maximum value for traffic at any given point for the heatmap
          if self._max_traffic < cumulative_traffic[link_id]:
            self._max_traffic = cumulative_traffic[link_id]
            self._max_traffic_link = link_id
            self._max_traffic_tick = self._tick
        elif event_type in ['left link', 'vehicle leaves traffic']:
          self._traffic_per_tick[self._tick][link_id] -= 1
          cumulative_traffic[link_id] -= 1
          ticked = True

        if ticked:
          self._tick += 1

  def _create_graph(self):
    for node_id, coords in self._nodes.items():
      self._G.add_node(node_id, pos=coords, is_facility=node_id.startswith('facility_'))
    for link in self._links:
      self._G.add_edge(link['from'], link['to'], id=link['id'])
    self._pos = nx.get_node_attributes(self._G, 'pos')

    if self._show_nearby_nodes:
      self._identify_close_nodes(radius=100) # TODO: move to load data

  def _identify_close_nodes(self, radius):
    facility_coords = {node: self._pos[node] for node in self._G.nodes if self._G.nodes[node].get('is_facility')}

    for node, coords in self._pos.items():
      if self._G.nodes[node].get('is_facility'):
        continue

      for facility_node, facility_coord in facility_coords.items():
        if self._euclidean_distance(coords, facility_coord) <= radius:
          self._close_to_facility_nodes.add(node)
          break

  def _setup_color_mapping(self):
    self.norm = colors.Normalize(vmin=0, vmax=self._max_traffic/2)

  def _update(self, frame):
    traffic_change = self._traffic_per_tick[self._tick]

    edge_colors_dict = {}
    edge_widths_dict = {}

    for link in self._links:
      for link_id, change in traffic_change.items():
        if link_id == link['id']:
          link['vehicles'] += change

      edge_colors_dict[link['id']] = self._cmap(self.norm(link['vehicles']))
      edge_widths_dict[link['id']] = 1 + self.norm(link['vehicles'])

    edgelist = [(link['from'], link['to']) for link in self._links]
    edge_colors = [edge_colors_dict[link["id"]] for link in self._links]
    edge_widths = [edge_widths_dict[link["id"]] for link in self._links]

    plt.cla()

    # Draw Facilities
    if self._show_facilities:
      facility_nodes = {node: pos for node, pos in self._pos.items() if self._G.nodes[node]['is_facility']}
      nx.draw_networkx_nodes(self._G, facility_nodes, facility_nodes.keys(), node_size=1, node_color='red')

    # Draw Network
    if self._show_nearby_nodes:
      close_nodes = {node: self._pos[node] for node in self._close_to_facility_nodes if node in self._pos}
      nx.draw_networkx_nodes(self._G, self._pos, close_nodes.keys(), node_size=1, node_color='blue')
    nx.draw_networkx_edges(self._G, self._pos, edgelist=edgelist, width=edge_widths, edge_color=edge_colors, edge_cmap=self._cmap)
    plt.title(f"Time: {self._tick}")
    self._tick += 1

  def visualize(self, fps=15, colormap='inferno'):
    self._load_data()
    self._create_graph()
    self._setup_color_mapping()

    fig, ax = plt.subplots()
    ax.set_aspect((self._max_x - self._min_x) / (self._max_y - self._min_y))

    if hasattr(cm, colormap):
      self._cmap = getattr(cm, colormap)
    else:
      print(f"Colormap '{colormap}' not recognized. Falling back to 'inferno'. Please select from: 'inferno', 'plasma', 'magma', 'spring', 'summer', 'autumn', or 'winter'.")
      self._cmap = cm.inferno

    sm = plt.cm.ScalarMappable(cmap=self._cmap, norm=self.norm)
    sm.set_array([])
    plt.colorbar(sm, ax=ax, label='Traffic Density')

    self._tick = 0
    ani = FuncAnimation(fig, self._update, frames=len(self._traffic_per_tick), repeat=False)
    ani.save(f"{self._output_file_path}/traffic_animation.gif", writer='ffmpeg', fps=fps)
    plt.show()

  def _euclidean_distance(self, coord1, coord2):
    return ((coord1[0] - coord2[0]) ** 2 + (coord1[1] - coord2[1]) ** 2) ** 0.5