forked from karinemiras/evoman_framework
-
Notifications
You must be signed in to change notification settings - Fork 0
/
competition_results_v2.py
178 lines (151 loc) · 7.08 KB
/
competition_results_v2.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
import sys
import matplotlib.pyplot as plt
from math import pi
from matplotlib.ticker import AutoLocator
from matplotlib.offsetbox import AnchoredText
sys.path.insert(0, 'evoman')
from environment import Environment
from demo_controller import player_controller
import numpy as np
import pandas as pd
import pickle as pkl
import os
import pygame
os.environ["SDL_VIDEODRIVER"] = "dummy"
###### CREATE A FOLDER CALLED solutions IN THE SAME DIRECTORY AS THIS SCRIPT AND PASTE ALL SOLUTION TXTs THERE ! #####
mode = "test" # Can be test for generating competition files, or demo to just present the winners
######
experiment_name = 'test'
n_enemies = 8
n_hidden = 10
# Switch for demo
if mode == "demo":
repetitions = 1
speed = "normal"
fullscreen = True
sound = "on"
else:
repetitions = 5
speed = "fastest"
fullscreen = False
sound = "off"
if not os.path.exists(experiment_name):
os.makedirs(experiment_name)
# Run each enemy n times for each group and record the data
df = pd.DataFrame(columns=["fitness", "player_life", "enemy_life", "time", "group", "repetition", "enemy"])
enemies = range(1, n_enemies + 1)
index = 0
for file in os.listdir("solutions"):
if file.endswith(".txt"):
group_name = file.replace(".txt", "")
try:
solution = np.loadtxt("solutions/" + file)
print("File of group " + str(group_name) + " was read")
except:
print("File of group "+str(group_name)+" could NOT be read")
for enemy in enemies:
env = Environment(
experiment_name=experiment_name,
enemies=[enemy],
playermode="ai",
fullscreen=fullscreen,
player_controller=player_controller(n_hidden),
enemymode="static",
level=2,
sound=sound,
speed=speed)
n_vars = (env.get_num_sensors() + 1) * n_hidden + (n_hidden + 1) * 5 # multilayer with 50 neurons
for n in range(repetitions):
try:
f, p, e, t = env.play(pcont=solution)
df.loc[index,] = [f, p, e, t, group_name, n, enemy]
index += 1
except:
print('bad solutioon')
if mode == "test":
# Convert time to time left for sorting
df["time"] = 3000 - df["time"]
df["gain"] = df["player_life"] - df["enemy_life"]
# Calculate gain and aggregate data
df_final = pd.DataFrame(columns=["group", "enemies_slain", "gain", "player_life", "enemy_life", "time"])
for i, group in enumerate(list(set(df["group"]))):
this_group = df["group"] == group
dead_enemies = np.count_nonzero(df["enemy_life"].loc[this_group] == 0) / repetitions
gain = sum(df["player_life"].loc[this_group] - df["enemy_life"].loc[this_group]) / repetitions
plife = sum(df["player_life"].loc[this_group]) / repetitions / n_enemies
elife = sum(df["enemy_life"].loc[this_group]) / repetitions / n_enemies
time = sum(df["time"].loc[this_group]) / repetitions / n_enemies
df_final.loc[i] = {"group": group, "enemies_slain": dead_enemies, "gain": gain, "player_life": plife, "enemy_life": elife, "time": time}
# Determine and print winners
winners = pd.DataFrame(columns=["slain", "gain"])
winners_slain = df_final.sort_values(by=["enemies_slain", "player_life", "time"], ascending=False).reset_index()
winners_gain = df_final.sort_values(by="gain", ascending=False).reset_index()
print("Winner for slain enemies: \n", winners["slain"].head(n=3))
print("Winner for gain measure: \n", winners["gain"].head(n=3))
# Index as ranks
winners_slain["time"] = 3000 - winners_slain["time"]
winners_gain["time"] = 3000 - winners_gain["time"]
pd.concat([winners_slain, winners_gain], axis=1).to_csv("winners.csv")
# Prepare data for radar chart and make plots of winners and whole class
# adapted from: https://python-graph-gallery.com/391-radar-chart-with-several-individuals/
for winner in [winners_slain["group"], winners_gain["group"], ["whole_class"]]: # !!!! [:3]
for group in winner:
this_group = df["group"] == group
if group == "whole_class":
df_plot = df.drop(["group", "repetition"], axis=1).apply(pd.to_numeric).groupby(
"enemy").mean().transpose()
else:
df_plot = df.loc[this_group].drop(["group", "repetition"], axis=1).apply(pd.to_numeric).groupby(
"enemy").mean().transpose()
df_plot = df_plot.reset_index().rename(columns={"index": "group"})
# Build radar chart
categories = list(df_plot)[1:]
N = len(categories)
# Determine angle
angles = [n / float(N) * 2 * pi for n in range(N)]
angles += angles[:1]
# Initialise radar chart
ax = plt.subplot(111, polar=True)
plt.title(group)
# If you want the first axis to be on top:
ax.set_theta_offset(pi / 2)
ax.set_theta_direction(-1)
# Draw one axe per variable + add labels labels yet
g = str(round(df_plot.drop("group", axis=1).loc[4].sum(), 2))
text_box = AnchoredText("Gain: " + g, frameon=False, loc=8, pad=-3.5)
plt.setp(text_box.patch, facecolor='white', alpha=0.5)
plt.gca().add_artist(text_box)
plt.xticks(angles[:-1], categories)
# Draw ylabels
ax.set_rlabel_position(0)
labels = ["gain"]
indices = [4]
for lab, col in zip(labels, indices):
values = df_plot.loc[col].drop('group').values.flatten().tolist()
values += values[:1]
ax.plot(angles, values, linewidth=1, linestyle='solid', label="energy")
ax.fill(angles, values, 'b', alpha=0.1)
if group == "whole_class" and lab == "gain":
print(lab, values)
ax.yaxis.set_major_locator(AutoLocator())
if lab == "player life":
continue
# Next line is to prevent that there is no plain in the plot when almost all values are 0 and one or two
# are really high
plt.ylim(bottom=min(values) - 10)
plt.legend(loc='lower right', bbox_to_anchor=(0.1, 0.1))
plt.savefig(group + "_energy.png", dpi=300)
plt.close()
ax = plt.subplot(111, polar=True)
ax.set_theta_offset(pi / 2)
ax.set_theta_direction(-1)
plt.xticks(angles[:-1], categories)
ax.set_rlabel_position(0)
if group != "whole_class":
plt.legend(loc='lower right', bbox_to_anchor=(0.1, 0.1))
plt.savefig(group + "_energy.png", dpi=300)
plt.close()
plt.close()
plt.hist(pd.to_numeric(df_final["gain"]))
plt.title("Distribution of gain\n(whole class)")
plt.savefig("gain_hist_whole_group.png", dpi=300)