added script for random vs greedy comp

utils/plot_results.py

@@ -2,12 +2,11 @@
 
 from sys import argv, exit
 import os
-from csv import writer
 from typing import Tuple, List
 import re
-import numpy as np
-from pprint import pprint
 
+import numpy as np
+import pandas as pd
 import matplotlib.pyplot as plt
 
 if len(argv) != 2:
@@ -18,8 +17,6 @@ path = argv[1]
 
 files = [f for f in os.listdir(path) if os.path.isfile(f"{ path }/{f}")]
 
-# files = [f for f in files if re.match(r"greedy_64_.+", f) is not None ]
-
 
 def parse_file(file: str) -> Tuple[float, List[int]]:
 
@@ -41,57 +38,91 @@ def parse_file(file: str) -> Tuple[float, List[int]]:
 
     return time, pops
 
-results = dict()
 
-with open("times.csv", "w+") as times_file:
+frames = list()
-    times = writer(times_file)
 
-    with open("pops.csv", "w+") as pops_file:
+for i, file in enumerate(files):
-        pops = writer(pops_file)
+    time, pops = parse_file(f"{ path }/{ file }")
 
-        for file in files:
+    parts = file.split(".")[0].split("_")
-            name = file.split(".")[0]
-            full_path = f"{ path }/{ file }"
-            time, pop = parse_file(full_path)
 
-            total_pops = sum(pop)/len(pop)
+    data = {
+            "name": parts[0],
+            "time": time,
+            "pops": int(sum(pops)),
+            }
 
-            results[name] = (total_pops, time)
+    if len(parts) > 1:
+        data["num_landmarks"] = int(parts[1])
+    if len(parts) > 2:
+        data["best_size"] = int(parts[2])
 
-            times.writerow([name, time])
+    frames.append(pd.DataFrame(data, index=[i]))
 
-            pops.writerow([name, *pop])
+results = pd.concat(frames)
-rel_pops = list()
-abs_pops = list()
-rel_time = list()
-abs_time = list()
-labels = list()
 
 
+# dijkstra vs A* vs greedy and random ALT
 
-pprint(results)
+#condition = ((results["name"] == "astar")
-baseline = "dijkstra"
+#        | (results["name"] == "dijkstra")
-base_pops = results[baseline][0]
+#        | ((results["num_landmarks"] == 32) & (results["best_size"] == 4)))
-base_time = results[baseline][1]
+#
-
+# basic = results[condition].sort_values(by=["time"], ascending=False)
-for name, values in results.items():
+# names = basic["name"].to_list()
-    pops, time = values
+# for i in range(len(names)):
-    labels.append(name)
+#     if names[i] == "greedy":
-    rel_pops.append(pops/base_pops)
+#         names[i] = "ALT 32 greedy"
-    rel_time.append(time/base_time)
+#     if names[i] == "random":
-    abs_pops.append(pops)
+#         names[i] = "ALT 32 random"
-    abs_time.append(time)
+# times = basic["time"] * 1000
+# 
+# fig, ax = plt.subplots()
+# 
+# bar = ax.bar(names, times)
+# ax.set_title("Average query time by algorithm")
+# ax.set_ylabel("time in ms")
+# ax.bar_label(bar)
+# fig.savefig("basic.svg")
+# plt.show()
 
 
+# comparison greedy vs. random, best_size = 4
 
-x = np.arange(len(labels))  # the label locations
+condition = (((results["name"] == "greedy") | (results["name"] == "random")) & (results["best_size"] == 4))
-width = 0.35  # the width of the bars
+
+data = results[condition]
+greedy = data[data["name"] == "greedy"]
+random = data[data["name"] == "random"]
+
+print(greedy)
+print(random)
+
+labels = greedy["num_landmarks"].to_list()
+labels = [int(label) for label in labels]
+greedy_times = greedy["time"].to_list()
+random_times = random["time"].to_list()
+
+width = 0.35
 
 fig, ax = plt.subplots()
-rects1 = ax.bar(x - width/2, abs_time , width, label='time')
-# rects2 = ax.bar(x + width/2, abs_pops, width, label='pops')
 
-ax.legend()
+x = np.arange(len(labels))
+
+greedy_bars = ax.bar(x - width/2, greedy_times, width, label="Greedy")
+random_bars = ax.bar(x + width/2, random_times, width, label="Random")
+ax.set_title("Comparison of ALT with random and greedy landmark selection")
+ax.set_ylabel("time in ms")
+ax.set_xlabel("number of landmarks")
 ax.set_xticks(x, labels)
+ax.legend()
+# ax.bar_label(greedy_bars)
+# ax.bar_label(random_bars)
+
+fig.save("random_greedy_comp.svg")
 
 plt.show()
+
+
+# comparison greedy best_size
+# comparison random best_size