.

intersection_switching/data_process_appendix.py

@@ -0,0 +1,243 @@
+import pickle
+import numpy as np
+import matplotlib.pyplot as plt
+import glob
+from radar_plot import ComplexRadar 
+
+low_balanced = [11, 11]
+low_unbalanced = [11, 6]
+
+medium_balanced = [22, 22]
+medium_unbalanced = [22, 11]
+
+high_balanced = [32, 32]
+high_unbalanced = [32, 16]
+
+
+# traffic_conditions = [low_balanced, low_unbalanced, medium_balanced, medium_unbalanced, high_balanced, high_unbalanced]
+
+traffic_dict = {('low', 'balanced'): [11,11],
+                      ('low', 'unbalanced'): [11,6],
+                      ('medium', 'balanced'): [22,22],
+                      ('medium', 'unbalanced'): [22,11],
+                      ('high', 'balanced'): [32,32],
+                      ('high', 'unbalanced'): [32,16],
+                     }
+
+pref_types = ['speed', 'stops', 'wait']
+
+vote_speed = [1.0, 0.0, 0.0]
+vote_stops = [0.0, 1.0, 0.0]
+vote_wait = [0.0, 0.0, 1.0]
+
+vote_uniform_1 = [0.5, 0.5, 0.0]
+vote_uniform_2 = [0.5, 0.0, 0.5]
+vote_uniform_3 = [0.0, 0.5, 0.5]
+
+vote_quarter_1 = [0.75, 0.25, 0.0]
+vote_quarter_2 = [0.75, 0.0, 0.25]
+vote_quarter_3 = [0.25, 0.0, 0.75]
+vote_quarter_4 = [0.0, 0.25, 0.75]
+vote_quarter_5 = [0.25, 0.75, 0.0]
+vote_quarter_6 = [0.0, 0.75, 0.25]
+
+# vote_types = [vote_speed, vote_stops, vote_wait, vote_uniform_1, vote_uniform_2, vote_uniform_3]#, vote_quarter_1, vote_quarter_2, vote_quarter_3, vote_quarter_4, vote_quarter_5, vote_quarter_6]
+
+# vote_types = [vote_uniform_1, vote_uniform_2, vote_uniform_3]
+vote_types = [vote_stops, vote_wait, vote_uniform_3]
+
+categories = ['Speed', 'Number of Stops', 'Wait Time']
+categories = [*categories, categories[0]]
+vote_types = [vote_uniform_3, ['both']]
+
+
+# all_data = {}
+# all_names = {}
+
+# for j, (label, traffic) in enumerate(traffic_dict.items()):
+#     data = []
+#     names = []
+#     for folder in ['proportional_100','both_lin_combo_test', 'both_cobb_doug_test']:
+#         if 'both' in folder:
+#             vote = ['both']
+#         else:
+#             vote = vote_uniform_3
+#         avg_total_waits = []
+#         avg_total_stops = []
+#         avg_total_speeds = []
+
+
+#         _path = f"../{folder}/{'_'.join(map(str,traffic))}_{'_'.join(map(str,vote))}"
+
+#         # for trial, path in enumerate([_path]):
+#         for trial, path in enumerate(glob.glob(f'{_path}*')):
+#             print('globbing', path)
+#             speeds_path = path + "/veh_speed_hist.pickle"
+#             stops_path = path + "/veh_stops.pickle"
+#             wait_path = path + "/veh_wait_time.pickle"
+
+#             with open(speeds_path, "rb") as f:
+#                 speeds = pickle.load(f)
+
+#             avg_speed = np.mean([np.mean(speeds[x]) for x in speeds.keys()])
+#             var_speed = np.std(([np.mean(speeds[x]) for x in speeds.keys()]))
+
+#             with open(stops_path, "rb") as f:
+#                 stops = pickle.load(f)
+
+#             avg_stops = np.mean([np.mean(stops[x]) for x in stops.keys()])
+#             var_stops = np.std(([np.mean(stops[x]) for x in stops.keys()]))
+
+#             with open(wait_path, "rb") as f:
+#                 wait = pickle.load(f)
+
+#             wait = {k:v if v else [0] for k,v in wait.items()}
+
+#             avg_wait = np.mean([np.mean(wait[x]) for x in wait.keys()])
+#             var_wait = np.std(([np.mean(wait[x]) for x in wait.keys()]))
+
+#             avg_total_waits.append(avg_wait)
+#             avg_total_stops.append(avg_stops)
+#             avg_total_speeds.append(avg_speed)
+
+#         result = [np.mean(avg_total_speeds), np.mean(avg_total_stops), np.mean(avg_total_waits)]
+#         result = [*result, result[0]]
+#         data.append(result)
+
+#         if vote == vote_uniform_3:
+#             name = "Prop S+W"
+#         elif vote == ['both']:
+#             if folder == 'both_lin_combo_test':
+#                 name = "Linear comb."
+#             elif folder == 'both_cobb_doug_test':
+#                 name = "Cobb Doug"
+
+#         names.append(name)
+
+
+#     key = f"{'_'.join(map(str,traffic))}"
+#     all_data.update({key:data})
+#     all_names.update({key:names})
+
+#     variables = ('Speed', 'Stops', 'Wait Time')
+#     ranges = [(0, max([x[0] for x in data])), (max([x[1] for x in data]), 0), (max([x[2] for x in data]), 0)]            
+#     # plotting
+
+
+
+# data = {'all_data': all_data, 
+#         'all_names': all_names}
+
+# with open('data.pickle', 'wb') as f:
+#     pickle.dump(data, f)
+
+with open('data.pickle', "rb") as f:
+    data = pickle.load(f)
+all_data = data['all_data']
+all_names = data['all_names']
+
+plt.rcParams.update({'font.size': 12})
+b1,b2, b3 = 0,0,0
+
+for label, traffic in traffic_dict.items():
+
+    _labels = [f"{'_'.join(map(str,traffic))}"]
+
+    b1 = max(b1, max([x[0] for ll in _labels for x in all_data[ll] ]))
+    b2 = max(b2, max([x[1] for ll in _labels for x in all_data[ll] ]))
+    b3 = max(b3, max([x[2] for ll in _labels for x in all_data[ll] ]))
+ranges = [(0, b1),
+            (b2, 0),
+            (b3, 0)]
+
+# for key, data in all_data.items():
+for label, traffic in traffic_dict.items():
+    _labels = f"{'_'.join(map(str,traffic))}"
+
+    fig1, axes = plt.subplots(1,1, subplot_kw={'projection':'polar'})
+
+    # for name in ["Prop S+W", "Linear comb.", "Cobb Doug"]:
+    key = f"{'_'.join(map(str,traffic))}"
+    data = all_data[key]
+    names = all_names[key]
+
+    b1 = max([x[0] for x in data ])
+    b2 = max([x[1] for x in data ])
+    b3 = max([x[2] for x in data ])
+    ranges = [(0, b1),
+                (b2, 0),
+                (b3, 0)]
+
+    save_name = f"../figs/{'_'.join(map(str,traffic))}_appendix.pdf"
+    print(save_name, ranges)
+    print(data)
+    # save_name = f"../figs/figure1.pdf"
+
+
+    variables = ('Speed', 'Stops', 'Wait Time')
+    radar = ComplexRadar(axes, variables, ranges)
+
+    for d, name in zip(data, names):
+        radar.plot(d, label=name)
+        radar.fill(d, alpha=0.2)
+
+    fig1.legend()
+
+    # if traffic[0] == 11:
+    #     if traffic[1] == 11:
+    #         axes.set_title("Low Balanced")
+    #     else:
+    #         axes.set_title("Low Unbalanced")
+    # elif traffic[0] == 22:
+    #     if traffic[1] == 22:
+    #         axes.set_title("Medium Balanced")
+    #     else:
+    #         axes.set_title("Medium Unbalanced")
+    # elif traffic[0] == 32:
+    #     if traffic[1] == 32:
+    #         axes.set_title("High Balanced")
+    #     else:
+    #         axes.set_title("High Unbalanced")
+    axes.set_title(' '.join(label))
+
+
+    fig1.savefig(save_name, format='pdf', bbox_inches='tight')
+
+    #####################
+
+
+#     fig1, axes = plt.subplots(1,1, subplot_kw={'projection':'polar'})
+
+#     radar = ComplexRadar(axes, variables, ranges)
+
+#     for d, name in zip(data, names):
+#         radar.plot(d, label=name)
+
+#         radar.fill(d, alpha=0.2)
+
+
+#     save_name = f"../figs/{traffic[0]}_{traffic[1]}.pdf"
+#     # save_name = f"../figs/figure1.pdf"
+#     fig1.legend()
+
+#     if traffic[0] == 11:
+#         if traffic[1] == 11:
+#             axes.set_title("Low Balanced")
+#         else:
+#             axes.set_title("Low Unbalanced")
+#     elif traffic[0] == 22:
+#         if traffic[1] == 22:
+#             axes.set_title("Medium Balanced")
+#         else:
+#             axes.set_title("Medium Unbalanced")
+#     elif traffic[0] == 32:
+#         if traffic[1] == 32:
+#             axes.set_title("High Balanced")
+#         else:
+#             axes.set_title("High Unbalanced")
+
+
+
+#     fig1.savefig(save_name, format='pdf', bbox_inches='tight')
+
+