-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathttesting.py
320 lines (296 loc) · 17.8 KB
/
ttesting.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
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
from itertools import combinations
from statsmodels.stats.weightstats import ttest_ind
import pandas as pd
import plotly.graph_objects as go
from sqlalchemy import create_engine
import numpy
engine = create_engine("mysql+pymysql://root:robotics4099@localhost/internship data")
connection = engine.connect()
count = 0
def increment():
global count
count = count+1
def plot_epochs(models, titles=[], legend=[],l=False):
d = []
for i in range(len(models)):
d.append(go.Scatter(name=f'{legend[i]} Player', x=[i for i in range(1, len(models[i][0]) + 1)],
y=[i for i in models[i][0]]))
d.append(go.Scatter(name=f'{legend[i]} Opponent', x=[i for i in range(1, len(models[i][1]) + 1)],
y=[i for i in models[i][1]]))
fig = go.Figure(data=d)
if l:
fig.update_yaxes(type="log")
fig.update_layout(template='none')
fig.update_layout(
title=titles[0],
xaxis_title='Epoch',
yaxis_title=titles[1],
legend_title='Model',
)
fig.show()
fig.write_image(f'./images/fig{count}.png')
increment()
def plot_mean_vals(vals_set, title_set, x_axis_set, y_axis_set, legend_set, calculate_mean=True, ALPHA=.05):
for vals, title, x_axis, y_axis, legend in zip(vals_set, title_set, x_axis_set, y_axis_set, legend_set):
list_vals = vals
if calculate_mean:
vals = {k: [sum(v[0]) / len(v[0]), sum(v[1]) / len(v[1]), numpy.std(v[0]), numpy.std(v[1])] for k, v in vals.items()}
sorted_vals = {k: v for k, v in sorted(vals.items(), key=lambda item: item[1][0], reverse=True)}
fig = go.Figure(data=[
go.Bar(name='Players', x=list(sorted_vals.keys()), y=[i[0] for i in sorted_vals.values()],error_y=dict(type='data',array=[i[2] for i in sorted_vals.values()])),
go.Bar(name='Opponents', x=list(sorted_vals.keys()), y=[i[1] for i in sorted_vals.values()],error_y=dict(type='data',array=[i[3] for i in sorted_vals.values()]))
])
fig.update_layout(barmode='group')
fig.update_layout(template='none')
fig.update_layout(yaxis=dict(range=[0, 1]))
fig.update_yaxes(nticks=40)
fig.update_layout(
title=title,
xaxis_title=x_axis,
yaxis_title=y_axis,
legend_title=legend,
)
fig.show()
fig.write_image(f'./images/fig{count}.png')
increment()
for combo in combinations(list(list_vals.keys()), 2):
_, p_val, _ = ttest_ind(list_vals[combo[0]][0], list_vals[combo[1]][0], usevar='unequal')
if p_val < ALPHA / len(list_vals):
_, high_p_val, _ = ttest_ind(list_vals[combo[0]][0], list_vals[combo[1]][0], usevar='unequal',
alternative='larger')
_, low_p_val, _ = ttest_ind(list_vals[combo[0]][0], list_vals[combo[1]][0], usevar='unequal',
alternative='smaller')
if high_p_val < ALPHA:
print(f'{combo[0]} Player is better than {combo[1]} Player')
elif low_p_val < ALPHA:
print(f'{combo[1]} Player is better than {combo[0]} Player')
else:
print(f'{combo[0]} Player and {combo[1]} Player are not equal')
else:
print(f'{combo[0]} Player and {combo[1]} Player cannot be proven to be not equal')
for combo in combinations(list(list_vals.keys()), 2):
_, p_val, _ = ttest_ind(list_vals[combo[0]][1], list_vals[combo[1]][1], usevar='unequal')
if p_val < ALPHA / len(list_vals):
_, high_p_val, _ = ttest_ind(list_vals[combo[0]][1], list_vals[combo[1]][1], usevar='unequal',
alternative='larger')
_, low_p_val, _ = ttest_ind(list_vals[combo[0]][1], list_vals[combo[1]][1], usevar='unequal',
alternative='smaller')
if high_p_val < ALPHA:
print(f'{combo[0]} Opponent is better than {combo[1]} Opponent')
elif low_p_val < ALPHA:
print(f'{combo[1]} Opponent is better than {combo[0]} Opponent')
else:
print(f'{combo[0]} Opponent and {combo[1]} Opponent are not equal')
else:
print(f'{combo[0]} Opponent and {combo[1]} Opponent cannot be proven to be not equal')
print()
# TQL 3 Red: saved_data/GAMMA=0.3 EPSILON=0.5 ALPHA=0.38 EPSILON_DECAY=0.99997Red,2020 08 24-08 35 45.pkl.xz
TQL3R = pd.read_pickle(
"./saved_data/GAMMA=0.3 EPSILON=0.5 ALPHA=0.38 EPSILON_DECAY=0.99997Red,2020 08 24-08 35 45.pkl.xz")
# TQL 3 Blue: saved_data/GAMMA=0.3 EPSILON=0.5 ALPHA=0.38 EPSILON_DECAY=0.99997Blue,2020 08 24-08 35 47.pkl.xz
TQL3B = pd.read_pickle(
"./saved_data/GAMMA=0.3 EPSILON=0.5 ALPHA=0.38 EPSILON_DECAY=0.99997Blue,2020 08 24-08 35 47.pkl.xz")
# DQN 3 Blue: saved_data/HIDDEN_LAYER_SIZE=116 BUFFER_SIZE=182 BATCH_SIZE=72 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0073632 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Blue,2020 08 25-09 55 51.pkl.xz
DQN3B = pd.read_pickle(
"./saved_data/HIDDEN_LAYER_SIZE=116 BUFFER_SIZE=182 BATCH_SIZE=72 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0073632 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Blue,2020 08 25-09 55 51.pkl.xz")
# DQN 3 Red: saved_data/HIDDEN_LAYER_SIZE=116 BUFFER_SIZE=182 BATCH_SIZE=72 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0073632 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 25-09 55 48.pkl.xz
DQN3R = pd.read_pickle(
"./saved_data/HIDDEN_LAYER_SIZE=116 BUFFER_SIZE=182 BATCH_SIZE=72 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0073632 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 25-09 55 48.pkl.xz")
# GQN 3 Red: saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 26-09 26 39.pkl.xz
GQN3R = pd.read_pickle(
"./saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 26-09 26 39.pkl.xz")
# GQN 3 Blue: saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Blue,2020 08 26-09 26 39.pkl.xz
GQN3B = pd.read_pickle(
"./saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Blue,2020 08 26-09 26 39.pkl.xz")
# MCTS 3 Red: saved_data/Trials=200 C=0.01 EPSILON=0.5Red,2020 08 25-22 00 00.pkl.xz
MCTS3R = pd.read_pickle("./saved_data/Trials=200 C=0.01 EPSILON=0.5Red,2020 08 25-22 00 00.pkl.xz")
# MCTS 3 Blue: saved_data/Trials=200 C=0.01 EPSILON=0.5Blue,2020 08 25-22 00 00.pkl.xz
MCTS3B = pd.read_pickle("./saved_data/Trials=200 C=0.01 EPSILON=0.5Blue,2020 08 25-22 00 00.pkl.xz")
# GQN 3-3 Red: saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 30-03 20 27.pkl.xz
GQN33R = pd.read_pickle(
"./saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 30-03 20 27.pkl.xz")
# GQN 3-3 Blue: saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Blue,2020 08 30-03 20 28.pkl.xz
GQN33B = pd.read_pickle(
"./saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Blue,2020 08 30-03 20 28.pkl.xz")
# GQN 2-3 Red: saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 29-23 39 08.pkl.xz
GQN23R = pd.read_pickle(
"./saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 29-23 39 08.pkl.xz")
# GQN 2-3 Blue: saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Blue,2020 08 29-23 39 08.pkl.xz
GQN23B = pd.read_pickle(
"./saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Blue,2020 08 29-23 39 08.pkl.xz")
# DQNvMCTS Red: saved_data/HIDDEN_LAYER_SIZE=116 BUFFER_SIZE=182 BATCH_SIZE=72 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0073632 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 29-02 56 39.pkl.xz
DQNM3R = pd.read_pickle(
"saved_data/HIDDEN_LAYER_SIZE=116 BUFFER_SIZE=182 BATCH_SIZE=72 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0073632 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 29-02 56 39.pkl.xz")
# DQNvMCTs Blue: saved_data/Trials=200 C=0.01 EPSILON=0.5Blue,2020 08 29-02 56 39.pkl.xz
DQNM3B = pd.read_pickle("./saved_data/Trials=200 C=0.01 EPSILON=0.5Blue,2020 08 29-02 56 39.pkl.xz")
# TQL 4 Red: saved_data/GAMMA=0.3 EPSILON=0.5 ALPHA=0.38 EPSILON_DECAY=0.99997Red,2020 08 29-09 25 25.pkl.xz
TQL4R = pd.read_pickle(
"./saved_data/GAMMA=0.3 EPSILON=0.5 ALPHA=0.38 EPSILON_DECAY=0.99997Red,2020 08 29-09 25 25.pkl.xz")
# TQL 4 Blue: saved_data/GAMMA=0.3 EPSILON=0.5 ALPHA=0.38 EPSILON_DECAY=0.99997Blue,2020 08 29-09 25 43.pkl.xz
TQL4B = pd.read_pickle(
"./saved_data/GAMMA=0.3 EPSILON=0.5 ALPHA=0.38 EPSILON_DECAY=0.99997Blue,2020 08 29-09 25 43.pkl.xz")
# DQN 4 Red: saved_data/HIDDEN_LAYER_SIZE=116 BUFFER_SIZE=182 BATCH_SIZE=72 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0073632 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 27-17 34 27.pkl.xz
DQN4R = pd.read_pickle(
"./saved_data/HIDDEN_LAYER_SIZE=116 BUFFER_SIZE=182 BATCH_SIZE=72 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0073632 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 27-17 34 27.pkl.xz")
# DQN 4 Blue: saved_data/HIDDEN_LAYER_SIZE=116 BUFFER_SIZE=182 BATCH_SIZE=72 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0073632 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Blue,2020 08 27-17 34 28.pkl.xz
DQN4B = pd.read_pickle(
"./saved_data/HIDDEN_LAYER_SIZE=116 BUFFER_SIZE=182 BATCH_SIZE=72 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0073632 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Blue,2020 08 27-17 34 28.pkl.xz")
# GQN 4 Red: saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 28-10 59 27.pkl.xz
GQN4R = pd.read_pickle(
"./saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Red,2020 08 28-10 59 27.pkl.xz")
# GQN 4 Blue: saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Blue,2020 08 28-10 59 27.pkl.xz
GQN4B = pd.read_pickle(
"./saved_data/HIDDEN_LAYER_SIZE=34 BUFFER_SIZE=186 BATCH_SIZE=98 TARGET_MODEL_SYNC=8 LEARNING_RATE=0.0052994 GAMMA=0.3 EPSILON=0.5 EPSILON_DECAY=0.99997Blue,2020 08 28-10 59 27.pkl.xz")
# MCTS 4 Red: saved_data/Trials=200 C=0.01 EPSILON=0.5Red,2020 08 29-01 12 45.pkl.xz
MCTS4R = pd.read_pickle("./saved_data/Trials=200 C=0.01 EPSILON=0.5Red,2020 08 29-01 12 45.pkl.xz")
# MCTS 4 blue: saved_data/Trials=200 C=0.01 EPSILON=0.5Blue,2020 08 29-01 12 45.pkl.xz
MCTS4B = pd.read_pickle("./saved_data/Trials=200 C=0.01 EPSILON=0.5Blue,2020 08 29-01 12 45.pkl.xz")
# values = [
# {
# 'TQL 3': [[0.592, 0.624, 0.574], [0.414, 0.412, 0.376]],
# 'DQN 3': [[0.436, 0.484, 0.462], [0.396, 0.378, 0.396]],
# 'GQN-1 3': [[0.61, 0.616, 0.608], [0.414, 0.418, 0.42]],
# 'MCTS 3': [[0.998, 0.992, 0.958], [0.95, 0.948]],
# 'GQN-2 3': [[0.654, 0.626, 0.628], [0.372, 0.322, 0.336]],
# 'GQN-3 3': [[0.704, 0.712, 0.708], [0.372, 0.404, 0.374]],
# },
# {
# 'TQL 4': [[0.524, 0.512, 0.492], [0.454, 0.45, 0.486]],
# 'DQN 4': [[0.67, 0.608, 0.624], [0.606, 0.596, 0.588]],
# 'GQN 4': [[0.392, 0.36, 0.358], [0.408, 0.434, 0.432]],
# 'MCTS 4': [[.4, 0.4, 0.5], [.5, 0.6, 0.5]]
# }
# ]
# plot_mean_vals(values, ['Fig. 1: Win Rate of Models against a random player on R(3)', 'Fig. 2: Win Rate of Models against a random player on R(4)'], ['R(3) Models', 'R(4) Models'],
# ['% Win Rate', '% Win Rate'], ['Model Type', 'Model Type'])
#
# # # Win Rate
# plot_epochs([
# [GQN3R.loc[1,:], GQN3B.loc[1,:]],
# [MCTS3R.loc[1,:], MCTS3B.loc[1,:]],
# [GQN23R.loc[1,:], GQN23B.loc[1,:]],
# [GQN33R.loc[1,:], GQN33B.loc[1,:]],
# ],
# ['Fig. 3: Win Rate of GQN and MCTS Models on R(3)', 'Win Rate'], ["GQN 1","MCTS","GQN 2", "GQN 3"])
#
# plot_epochs([
#
# [GQN4R.loc[1,:], GQN4B.loc[1,:]],
# [MCTS4R.loc[1,:], MCTS4B.loc[1,:]],
#
# ],
# ['Fig. 4: Win Rate of GQN and MCTS Models on R(4)', 'Win Rate'], ["GQN 1","MCTS"])
# plot_epochs([
# [TQL3R.loc[1,:], TQL3B.loc[1,:]],
# [DQN3R.loc[1,:], DQN3B.loc[1,:]],
#
# ],
# ['Fig. 5 : Win Rate of TQL and DQN Models on R(3)', 'Win Rate'], ['TQL','DQN'])
#
# plot_epochs([
# [TQL4R.loc[1,:], TQL4B.loc[1,:]],
# [DQN4R.loc[1,:], DQN4B.loc[1,:]],
#
#
# ],
# ['Fig. 6: Win Rate of TQL and DQN Models on R(4)', 'Win Rate'], ['TQL','DQN'])
# # # Move Time
# plot_epochs([
#
# [GQN3R.loc[2,:], GQN3B.loc[2,:]],
# [MCTS3R.loc[2,:], MCTS3B.loc[2,:]],
# [GQN23R.loc[2,:], GQN23B.loc[2,:]],
# [GQN33R.loc[2,:], GQN33B.loc[2,:]],
# ],
# ['Fig. 7: Move Time of GQN and MCTS Models on R(3)', 'Move Time'], ["GQN 1","MCTS","GQN 2", "GQN 3"])
#
# plot_epochs([
#
# [GQN4R.loc[2,:], GQN4B.loc[2,:]],
# [MCTS4R.loc[2,:], MCTS4B.loc[2,:]],
#
# ],
# ['Fig. 8: Move Time of GQN and MCTS Models on R(4)', 'Move Time'], ["GQN 1","MCTS"])
#
# plot_epochs([
# [TQL3R.loc[2,:], TQL3B.loc[2,:]],
# [DQN3R.loc[2,:], DQN3B.loc[2,:]],
#
# ],
# ['Fig. 9: Move Time of TQL and DQN Models on R(3)', 'Move Time'], ['TQL','DQN'])
#
# plot_epochs([
# [TQL4R.loc[2,:], TQL4B.loc[2,:]],
# [DQN4R.loc[2,:], DQN4B.loc[2,:]],
#
#
# ],
# ['Fig. 10: Move Time of TQL and DQN Models on R(4)', 'Move Time'], ['TQL','DQN'])
#
# #Average Number of Moves
# plot_epochs([
# [GQN3R.loc[3, :].rolling(100,min_periods=0).mean(), GQN3B.loc[3, :].rolling(100,min_periods=0).mean()],
# [MCTS3R.loc[3, :].rolling(100,min_periods=0).mean(), MCTS3B.loc[3, :].rolling(100,min_periods=0).mean()],
# [GQN23R.loc[3, :].rolling(100,min_periods=0).mean(), GQN23B.loc[3, :].rolling(100,min_periods=0).mean()],
# [GQN33R.loc[3, :].rolling(100,min_periods=0).mean(), GQN33B.loc[3, :].rolling(100,min_periods=0).mean()],
#
# ],
# ['Fig. 11: Average Number of Moves of GQN and MCTS Models on R(3)', 'Average Number of Moves'],
# [ "GQN 1","MCTS","GQN 2", "GQN 3"])
# #
# plot_epochs([
# [GQN4R.loc[3, :].rolling(100,min_periods=0).mean(), GQN4B.loc[3, :].rolling(100,min_periods=0).mean()],
# [MCTS4R.loc[3, :].rolling(100,min_periods=0).mean(), MCTS4B.loc[3, :].rolling(100,min_periods=0).mean()],
#
# #
# ],
# ['Fig. 12: Average Number of Moves of GQN and MCTS Models on R(4)', 'Average Number of Moves'], [ "GQN 1","MCTS"])
#
# plot_epochs([
#
# [TQL3R.loc[3, :].rolling(100,min_periods=0).mean(), TQL3B.loc[3, :].rolling(100,min_periods=0).mean()],
# [DQN3R.loc[3, :].rolling(100,min_periods=0).mean(), DQN3B.loc[3, :].rolling(100,min_periods=0).mean()],
#
# ],
# ['Fig. 13: Average Number of Moves of TQL and DQN Models on R(3)', 'Average Number of Moves'],
# [ 'TQL','DQN',])
# #
# plot_epochs([
#
# [TQL4R.loc[3, :].rolling(100,min_periods=0).mean(), TQL4B.loc[3, :].rolling(100,min_periods=0).mean()],
# [DQN4R.loc[3, :].rolling(100,min_periods=0).mean(), DQN4B.loc[3, :].rolling(100,min_periods=0).mean()],
# #
# ],
# ['Fig. 14: Average Number of Moves of TQL and DQN Models on R(4)', 'Average Number of Moves'], [ 'TQL','DQN',],True)
#
# # Memory Usage
# plot_epochs([
# [GQN3R.loc[4, :], GQN3B.loc[4, :]],
# [MCTS3R.loc[4, :], MCTS3B.loc[4, :]],
# [GQN23R.loc[4, :], GQN23B.loc[4, :]],
# [GQN3R.loc[4, :], GQN3B.loc[4, :]],
# ],
# ['Fig. 15: Memory Usage of GQN and MCTS Models on R(3)', 'Memory Usage'], ["GQN 1", "MCTS", "GQN 2", "GQN 3"],True)
# plot_epochs([
# [GQN4R.loc[4, :], GQN4B.loc[4, :]],
# [MCTS4R.loc[4, :], MCTS4B.loc[4, :]],
# ],
# ['Fig. 16: Memory Usage of GQN and MCTS Models on R(4)', 'Memory Usage'], ["GQN 1", "MCTS"])
plot_epochs([
[TQL3R.loc[4, :], TQL3B.loc[4, :]],
[DQN3R.loc[4, :], DQN3B.loc[4, :]],
],
['Fig. 17: Memory Usage of TQL and DQN Models on R(3)', 'Memory Usage'], ['TQL','DQN'])
plot_epochs([
[TQL4R.loc[4, :], TQL4B.loc[4, :]],
[DQN4R.loc[4, :], DQN4B.loc[4, :]],
],
['Fig 18: Memory Usage of TQL and DQN Models on R(4)', 'Memory Usage'], ['TQL','DQN'])
# plot_mean_vals(values, ['% Win Rate of Models on R(3)', '% Win Rate of Models on R(4)'], ['R(3) Models', 'R(4) Models'],
# ['% Win Rate', '% Win Rate'], ['Model Type', 'Model Type'])
# names = ['TQL 3 Player','TQL 3 Opponent','DQN 3 Player','DQN 3 Opponent',"GQN-1 3 Player","GQN-1 3 Opponent","MCTS 3 Player","MCTS 3 Opponent","GQN-2 3 Player","GQN-2 3 Opponent", "GQN-3 3 Player","GQN-3 3 Opponent",'TQL 4 Player','TQL 4 Opponent', 'DQN 4 Player','DQN 4 Opponent', "GQN-1 4 Player","GQN-1 4 Opponent", "MCTS 4 Player", "MCTS 4 Opponent"]
# for i,n in zip([TQL3R, TQL3B,DQN3R, DQN3B,GQN3R, GQN3B,MCTS3R, MCTS3B,GQN23R, GQN23B,GQN33R, GQN33B,TQL4R, TQL4B, DQN4R, DQN4B,GQN4R, GQN4B,MCTS4R, MCTS4B],names):
# i = i.transpose()
# i.columns = ['Loss','Win Rate', 'Average Move Time', "Number of Moves", "Memory Usage"]
# i.to_sql(n,con=connection, if_exists="replace")
# print(i)