-
Notifications
You must be signed in to change notification settings - Fork 0
/
result_plotter.py
506 lines (420 loc) · 25.1 KB
/
result_plotter.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
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
#%% This is made as a jupyter notebook, it is made to be ran in the VSCODE interactive mode.
# But it still can be ran as a python file.
from collections import defaultdict
from genericpath import isfile
from pathlib import Path
from tokenize import group
from typing import List
import matplotlib.pyplot as plt
import json
import pandas as pd
from environment import Environment
import numpy as np
plt.rcParams.update({'font.size': 18})
import os
import matplotlib.markers as mmarkers
import matplotlib.patches as mpatches
from matplotlib import cm
import torch
import numpy as np
import matplotlib.lines as mlines
import math
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
# Color Palette
cp = ["#e60049", "#0bb4ff", "#50e991", "#e6d800", "#9b19f5", "#ffa300", "#dc0ab4", "#b3d4ff", "#00bfa0"]
# cp2 = ['#fd7f6f', '#bd7ebe', 'ffb55a', '#7eb0d5', '#b2e061']
# ["#fd7f6f", "#7eb0d5", "#b2e061", "#bd7ebe", "#ffb55a", "#ffee65", "#beb9db", "#fdcce5", "#8bd3c7"]
#%%
def prepare_metrics_df(models: List):
# args = defaultdict(list)
args = pd.DataFrame()
metrics = defaultdict(list)
for model_path in models:
with open(Path('result', model_path, 'result_metrics.json')) as json_file:
data = json.load(json_file)
metrics['model'].append(model_path)
metrics['avg_generated_line'].append(data['avg_generated_line'])
metrics['mean_sat_group_od'].append(data['mean_sat_group_od'])
metrics['mean_sat_group_od_pct'].append(data['mean_sat_group_od_pct'])
metrics['mean_sat_od_by_group'].append(data['mean_sat_od_by_group'])
metrics['mean_sat_od_by_group_pct'].append(data['mean_sat_od_by_group_pct'])
metrics['group_gini'].append(data['group_gini'])
metrics['group_pct_gini'].append(data['group_pct_gini'])
metrics['group_pct_diff'].append(1-abs(data['mean_sat_od_by_group_pct'][0] - data['mean_sat_od_by_group_pct'][1]))
if 'mean_distance' in data:
metrics['mean_distance'].append(data['mean_distance'])
if 'mean_group_distance' in data:
metrics['mean_group_distance'].append(data['mean_group_distance'])
argpath = Path('result', model_path, 'args.txt')
if argpath.is_file():
with open(argpath) as json_file:
data = json.load(json_file)
data['model'] = model_path
args = pd.concat([args, pd.DataFrame(data, index=[0])])
# args = args.append(data, ignore_index=True)
# for k in data.keys():
# args[k].append(data[k])
# args.append(data[k])
df_metrics = pd.DataFrame(metrics)
# df_args = pd.DataFrame(args)
if args.shape[0] > 0:
return pd.merge(args, df_metrics, how='right', on='model')
else:
return df_metrics
def print_ci(df, col, model: str, z=1.96):
m = df[col].mean()
std = df[col].std()
se = std/math.sqrt(df.shape[0])
print(f"[{model} - {col}] - Mean: {m} +- {z * se} SE: {se}, CI:({m - z * se}, {m + z * se}), Sample Size: {df.shape[0]}")
return m
def print_stats(df, model:str):
od = print_ci(df, 'mean_sat_group_od_pct', model)
gini = print_ci(df, 'group_pct_gini', model)
lq = print_ci(df, 'lowest_quintile_sat_od_pct', model)
print('--------')
return od, gini, lq
def plot_bar(model_names, model_labels, model_colors, model_hatches,
metrics_df: pd.DataFrame, env_name: str, figsize=(10, 5),
legend_loc='best',
group_names=('1st quintile', '2nd', '3rd', '4th', '5th')):
# https://stackoverflow.com/questions/10369681/how-to-plot-bar-graphs-with-same-x-coordinates-side-by-side-dodged
# Width of a bar
width = 0.3
ind = np.arange(len(group_names))
xpos = ind
fig, ax = plt.subplots(figsize=figsize)
for i, model in enumerate(model_names):
results = metrics_df.loc[metrics_df['model'] == model].iloc[0]['mean_sat_od_by_group_pct']
# TODO change that
results = [r * 100 for r in results]
# position of the bar on the x axis
# xpos = ind if i == 0 else ind + width
ax.bar(xpos, results , width, label=model_labels[i], color=model_colors[i], hatch=model_hatches[i])
xpos = xpos + width
plt.xlabel('House Price Quintiles', fontsize=32)
plt.ylabel('% of total satisfied flows', fontsize=32)
plt.title(f"Benefits Distribution among Groups - {env_name}", fontsize=32)
plt.xticks(ind + width * 3 / 2, group_names)
ax.legend(loc=legend_loc)
fig.tight_layout()
return fig
def plot_lines(env: Environment, model_names, model_labels: List, model_colors: List, model_markers: List, metrics_df: pd.DataFrame, env_name, lines=None, figsize=(15, 10), legend_loc="lower right"):
fig, ax = plt.subplots(figsize=figsize)
im1 = ax.imshow(env.grid_groups, cm.get_cmap('viridis'), alpha=0.3)
labels = ['1st quintile', '2nd quintile', '3rd quintile', '4th quintile', '5th quintile']
values = (np.unique(env.grid_groups[~np.isnan(env.grid_groups)]))
grid_colors = [ im1.cmap(im1.norm(value)) for value in values]
patches = [ mpatches.Patch(color=grid_colors[i], label=labels[i] ) for i in range(len(labels)) ]
ax.legend(handles=patches, loc=legend_loc, prop={'size': 14})
ax.set_title(f'Generated Lines - {env_name}', fontsize=32)
if lines:
for i, l in enumerate(lines):
# Note here we reverse the dimensions because on scatter plots the horizontal axis is the x axis.
l_v = env.vector_to_grid(l).cpu()
label = model_labels[i]
# label = "_no_legend"
# if i == 0:
# label = "Generated Metro Lines"
ax.plot(l_v[1], l_v[0], f"-{model_markers[i]}", color=model_colors[i], label=label, alpha=0.8, markersize=12, linewidth=4)
else:
# for i, l in enumerate(metrics_df['avg_generated_line']):
for i, model in enumerate(model_names):
# for i, model in enumerate(metrics_df['model']):
with open(Path('result', model, 'tour_idx_multiple.txt')) as f:
l = [int(idx) for idx in f.readline().split(',')]
# l_v = env.vector_to_grid(torch.tensor(l).reshape(-1,1)).cpu()
l_v = env.vector_to_grid(torch.tensor([l])).T.cpu()
label = "_no_legend"
if i == 0:
label = "Generated Metro Lines"
ax.plot(l_v[:, 1], l_v[:, 0], f"-{model_markers[i]}", color=model_colors[i], label=label, alpha=1, markersize=12, linewidth=4)
fig.tight_layout()
return fig
# To create a scatterplot with different custom markers.
# From https://github.com/matplotlib/matplotlib/issues/11155
def plot_scatter(x,y,ax=None, markers=None, labels=None, colors=None, env_name=None, figsize=(12, 8)):
s = np.repeat(800, len(x))
fig, ax = plt.subplots(figsize=figsize)
# if not ax: ax = plt.gca()
sc = ax.scatter(x,y, s=s, c=colors)
if (markers is not None) and (len(markers)==len(x)):
paths = []
for marker in markers:
if isinstance(marker, mmarkers.MarkerStyle):
marker_obj = marker
else:
marker_obj = mmarkers.MarkerStyle(marker)
path = marker_obj.get_path().transformed(
marker_obj.get_transform())
paths.append(path)
sc.set_paths(paths)
ax.set_xlabel('% of total satisfied flows', fontsize=32)
ax.set_ylabel('Gini Index', fontsize=32)
fig.suptitle(f'Equity vs Efficiency - {env_name}', fontsize=32)
ax.set_ylim((0,0.8))
markers = [mlines.Line2D([], [], color=colors[i], marker=markers[i], linestyle='None',
markersize=10, label=labels[i]) for i in range(len(s))]
ax.legend(handles=markers, prop={'size': 24})
fig.tight_layout()
return fig
def create_all_plots(env: Environment, metrics_df: pd.DataFrame, metadata: List, bar_plot_models: List, line_plot_models: List, scatter_plot_models: List, scatter_x: List, scatter_y: List, plot_name_prefix: None, env_name=None, figsize=(12,8), group_names=('1st quintile', '2nd', '3rd', '4th', '5th')):
metadata = pd.DataFrame(metadata, columns=['label', 'model', 'color', 'pattern', 'marker'])
metadata.index = metadata['label']
bar_models = metadata.loc[bar_plot_models]
bar_fig = plot_bar(
bar_models['model'].tolist(),
bar_models['label'].tolist(),
bar_models['color'].tolist(),
bar_models['pattern'].tolist(),
metrics_df[metrics_df.index.isin(bar_models['model'].tolist())],
env_name=env_name,
figsize=figsize,
group_names=group_names)
bar_fig.savefig(f'./{plot_name_prefix}_bar.png')
line_models = metadata.loc[line_plot_models]
line_fig = plot_lines(
env,
line_models['model'].tolist(),
line_models['label'].tolist(),
line_models['color'].tolist(),
line_models['marker'].tolist(),
metrics_df[metrics_df.index.isin(line_models['model'].tolist())],
env_name=env_name,
legend_loc="lower left",
figsize=figsize)
line_fig.savefig(f'./{plot_name_prefix}_lines.png')
scatter_models = metadata.loc[scatter_plot_models]
scatter_fig = plot_scatter(
x=scatter_x,
y=scatter_y,
markers=scatter_models['marker'].tolist(),
labels=scatter_models['label'].tolist(),
colors=scatter_models['color'].tolist(),
env_name=env_name,
figsize=figsize)
scatter_fig.savefig(f'./{plot_name_prefix}_scatter.png')
# %% AMSTERDAM
amsterdam = Environment(Path(f"./environments/amsterdam/"), groups_file='price_groups_5.txt')
models = []
paths = [ f.path for f in os.scandir('./result') if f.is_dir() ]
for p in paths:
if 'amsterdam' in p:
models.append(p.split('/')[-1])
metrics_ams = prepare_metrics_df(models)
metrics_ams['mean_sat_group_od_pct'] = metrics_ams['mean_sat_group_od_pct'] * 100
metrics_ams['lowest_quintile_sat_od_pct'] = metrics_ams['mean_sat_od_by_group_pct'].str[0]
metrics_ams['lowest_quintile_sat_od_pct'] = metrics_ams['lowest_quintile_sat_od_pct'] * 100
metrics_ams.index = metrics_ams['model']
# %% Amsterdam Income Quintiles averages calculation
ams_empty_ses1 = metrics_ams[ ((metrics_ams['existing_lines'] == 0) | np.isnan(metrics_ams['existing_lines']))
& (pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'weighted')
& (metrics_ams['ses_weight'] == 1)
& (metrics_ams['var_lambda'] == 0)]
ams_empty_ses0 = metrics_ams[((metrics_ams['existing_lines'] == 0) | np.isnan(metrics_ams['existing_lines']))
& (pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'weighted')
& (metrics_ams['ses_weight'] == 0)
& (metrics_ams['var_lambda'] == 0)]
ams_empty_ggi_2 = metrics_ams[((metrics_ams['existing_lines'] == 0) | np.isnan(metrics_ams['existing_lines']))
& (pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'ggi')
& (metrics_ams['ggi_weight'] == 2)]
ams_empty_var_3 = metrics_ams[((metrics_ams['existing_lines'] == 0) | np.isnan(metrics_ams['existing_lines']))
& (pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'group')
& (metrics_ams['var_lambda'] == 3)]
ams_empty_rawls = metrics_ams[((metrics_ams['existing_lines'] == 0) | np.isnan(metrics_ams['existing_lines']))
& (pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'rawls')]
ams_full_ses1 = metrics_ams[ ((metrics_ams['existing_lines'] != 0) & ~np.isnan(metrics_ams['existing_lines']))
& (pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'weighted')
& (metrics_ams['ses_weight'] == 1)
& (metrics_ams['var_lambda'] == 0)]
ams_full_ses0 = metrics_ams[((metrics_ams['existing_lines'] != 0) & ~np.isnan(metrics_ams['existing_lines']))
& (pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'weighted')
& (metrics_ams['ses_weight'] == 0)
& (metrics_ams['var_lambda'] == 0)]
ams_full_var_3 = metrics_ams[((metrics_ams['existing_lines'] != 0) & ~np.isnan(metrics_ams['existing_lines']))
& (pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'group')
& (metrics_ams['var_lambda'] == 3)]
ams_full_ggi_2 = metrics_ams[((metrics_ams['existing_lines'] != 0) & ~np.isnan(metrics_ams['existing_lines']))
& (pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'ggi')
& (metrics_ams['ggi_weight'] == 2)]
ams_full_rawls = metrics_ams[((metrics_ams['existing_lines'] != 0) & ~np.isnan(metrics_ams['existing_lines']))
& (pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'rawls')]
ams_full_ses0_od, ams_full_ses0_gini, ams_full_ses0_lq = print_stats(ams_full_ses0, 'ams_full_ses_0')
ams_full_ses1_od, ams_full_ses1_gini, ams_full_ses1_lq = print_stats(ams_full_ses1, 'ams_full_ses_1')
ams_full_var_3_od, ams_full_var_3_gini, ams_full_var_3_lq = print_stats(ams_full_var_3, 'ams_full_var_3')
ams_full_rawls_od, ams_full_rawls_gini, ams_full_rawls_lq = print_stats(ams_full_rawls, 'ams_full_rawls')
ams_full_ggi_2_od, ams_full_ggi_2_gini, ams_full_ggi_2_lq = print_stats(ams_full_ggi_2, 'ams_full_ggi_2')
ams_empty_ses0_od, ams_empty_ses0_gini, ams_empty_ses0_lq = print_stats(ams_empty_ses0, 'ams_empty_ses_0')
ams_empty_ses1_od, ams_empty_ses1_gini, ams_empty_ses1_lq = print_stats(ams_empty_ses1, 'ams_empty_ses_1')
ams_empty_var_3_od, ams_empty_var_3_gini, ams_empty_var_3_lq = print_stats(ams_empty_var_3, 'ams_empty_var_3')
ams_empty_rawls_od, ams_empty_rawls_gini, ams_empty_rawls_lq = print_stats(ams_empty_rawls, 'ams_empty_rawls')
ams_empty_ggi_2_od, ams_empty_ggi_2_gini, ams_empty_ggi_2_lq = print_stats(ams_empty_ggi_2, 'ams_empty_ggi_2')
#%% Amsterdam Full environment
ams_full_plot = [
['Max. Effic.', 'amsterdam_20220810_09_33_35.507895', cp[0], '', 'o'],
['Access. Index', 'amsterdam_20220807_22_43_37.708173', cp[1], '-', 's'],
['Var.Reg', 'amsterdam_20220809_00_40_57.169847', cp[2], '+', '^'],
['Rawls', 'amsterdam_20220808_11_53_12.554688', cp[3], 'o', 'v'],
['GGI', 'amsterdam_20220810_20_23_40.289417', cp[4], '/', 'D'],
]
create_all_plots(amsterdam, metrics_ams, ams_full_plot,
bar_plot_models=['Max. Effic.', 'Rawls', 'GGI'],
line_plot_models=['Max. Effic.', 'Access. Index', 'Rawls', 'GGI', 'Var.Reg'],
scatter_plot_models=['Max. Effic.', 'Access. Index', 'Var.Reg', 'Rawls', 'GGI'],
scatter_x=[ams_full_ses0_od, ams_full_ses1_od, ams_full_var_3_od, ams_full_rawls_od, ams_full_ggi_2_od],
scatter_y=[ams_full_ses0_gini, ams_full_ses1_gini, ams_full_var_3_gini, ams_full_rawls_gini, ams_full_ggi_2_gini],
plot_name_prefix='ams_full',
env_name='Amsterdam')
ams_empty_plot = [
['Max. Effic.', 'amsterdam_20220705_18_17_31.196986', cp[0], '', 'o'],
['Access. Index', 'amsterdam_20220705_18_25_09.654804', cp[1], '-', 's'],
['Var.Reg', 'amsterdam_20220706_11_15_16.765435', cp[2], '+', '^'],
['Rawls', 'amsterdam_20220810_09_26_45.963603', cp[3], 'o', 'v'],
['GGI', 'amsterdam_20220708_11_21_23.191428', cp[4], '/', 'D'],
]
create_all_plots(amsterdam, metrics_ams, ams_empty_plot,
bar_plot_models=['Max. Effic.', 'Rawls', 'GGI'],
line_plot_models=['Max. Effic.', 'Access. Index', 'Rawls', 'GGI', 'Var.Reg'],
scatter_plot_models=['Max. Effic.', 'Access. Index', 'Var.Reg', 'Rawls', 'GGI'],
scatter_x=[ams_empty_ses0_od, ams_empty_ses1_od, ams_empty_var_3_od, ams_empty_rawls_od, ams_empty_ggi_2_od],
scatter_y=[ams_empty_ses0_gini, ams_empty_ses1_gini, ams_empty_var_3_gini, ams_empty_rawls_gini, ams_empty_ggi_2_gini],
plot_name_prefix='ams_empty',
env_name='Amsterdam')
#%% Amsterdam Weighted groups Experiment (Western-non-western)
ams_w_empty_ses1 = metrics_ams[ ((metrics_ams['existing_lines'] == 0) | np.isnan(metrics_ams['existing_lines']))
& (~pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'weighted')
& (metrics_ams['ses_weight'] == 1)
& (metrics_ams['var_lambda'] == 0)]
ams_w_empty_ses0 = metrics_ams[((metrics_ams['existing_lines'] == 0) | np.isnan(metrics_ams['existing_lines']))
& (~pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'weighted')
& (metrics_ams['ses_weight'] == 0)
& (metrics_ams['var_lambda'] == 0)]
ams_w_empty_ggi_2 = metrics_ams[((metrics_ams['existing_lines'] == 0) | np.isnan(metrics_ams['existing_lines']))
& (~pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'ggi')
& (metrics_ams['ggi_weight'] == 2)]
ams_w_empty_var_3 = metrics_ams[((metrics_ams['existing_lines'] == 0) | np.isnan(metrics_ams['existing_lines']))
& (~pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['reward'] == 'group')
& (metrics_ams['var_lambda'] == 3)]
ams_w_empty_rawls = metrics_ams[((metrics_ams['existing_lines'] == 0) | np.isnan(metrics_ams['existing_lines']))
& (~pd.isnull(metrics_ams['group_weights_files']))
& (metrics_ams['actor_lr'] == 15e-4)
& (metrics_ams['reward'] == 'rawls')]
ams_w_empty_ses1_od, ams_w_empty_ses1_gini, ams_w_empty_ses1_lq = print_stats(ams_w_empty_ses1, 'ams_w_empty_ses1')
ams_w_empty_ses0_od, ams_w_empty_ses0_gini, ams_w_empty_ses0_lq = print_stats(ams_w_empty_ses0, 'ams_w_empty_ses_0')
ams_w_empty_ggi_2_od, ams_w_empty_ggi_2_gini, ams_w_empty_ggi_2_lq = print_stats(ams_w_empty_ggi_2, 'ams_w_empty_ggi_2')
ams_w_empty_var_3_od, ams_w_empty_var_3_gini, ams_w_empty_var_3_lq = print_stats(ams_w_empty_var_3, 'ams_w_empty_var_3')
ams_w_empty_rawls_od, ams_w_empty_rawls_gini, ams_w_empty_rawls_lq = print_stats(ams_w_empty_rawls, 'ams_w_empty_rawls')
ams_w_empty_plot = [
['Max. Effic.', 'amsterdam_20221025_18_07_17.271867', cp[0], '', 'o'],
['Access. Index', 'amsterdam_20221026_16_31_11.370638', cp[1], '-', 's'],
['Var.Reg', 'amsterdam_20221026_18_17_07.310764', cp[2], '+', '^'],
['Rawls', 'amsterdam_20221027_09_51_21.915196', cp[3], 'o', 'v'],
['GGI', 'amsterdam_20221026_17_28_17.106600', cp[4], '/', 'D'],
]
create_all_plots(amsterdam, metrics_ams, ams_w_empty_plot,
bar_plot_models=['Max. Effic.', 'Rawls', 'GGI'],
line_plot_models=['Max. Effic.', 'Access. Index', 'Rawls', 'GGI', 'Var.Reg'],
scatter_plot_models=['Max. Effic.', 'Access. Index', 'Var.Reg', 'Rawls', 'GGI'],
scatter_x=[ams_w_empty_ses0_od, ams_w_empty_ses1_od, ams_w_empty_var_3_od, ams_w_empty_rawls_od, ams_w_empty_ggi_2_od],
scatter_y=[ams_w_empty_ses0_gini, ams_w_empty_ses1_gini, ams_w_empty_var_3_gini, ams_w_empty_rawls_gini, ams_w_empty_ggi_2_gini],
plot_name_prefix='ams_w_empty',
env_name='Amsterdam',
group_names=('Dutch/Western', 'Non-Western'))
#%% XIAN
xian = Environment(Path(f"./environments/xian/"), groups_file='price_groups_5.txt')
models = []
paths = [ f.path for f in os.scandir('./result') if f.is_dir() ]
for p in paths:
if 'xian' in p:
models.append(p.split('/')[-1])
metrics_xian = prepare_metrics_df(models)
metrics_xian['mean_sat_group_od_pct'] = metrics_xian['mean_sat_group_od_pct'] * 100
metrics_xian['lowest_quintile_sat_od_pct'] = metrics_xian['mean_sat_od_by_group_pct'].str[0]
metrics_ams['lowest_quintile_sat_od_pct'] = metrics_ams['lowest_quintile_sat_od_pct'] * 100
metrics_xian.index = metrics_xian['model']
xian_full_ses0 = metrics_xian[((metrics_xian['existing_lines'] != 0) & ~np.isnan(metrics_xian['existing_lines']))
& (metrics_xian['reward'] == 'weighted')
& (metrics_xian['ses_weight'] == 0)
& (metrics_xian['var_lambda'] == 0)]
xian_full_ses1 = metrics_xian[((metrics_xian['existing_lines'] != 0) & ~np.isnan(metrics_xian['existing_lines']))
& (metrics_xian['reward'] == 'weighted')
& (metrics_xian['ses_weight'] == 1)
& (metrics_xian['var_lambda'] == 0)]
xian_full_var_5 = metrics_xian[((metrics_xian['existing_lines'] != 0) & ~np.isnan(metrics_xian['existing_lines']))
& (metrics_xian['reward'] == 'group')
& (metrics_xian['var_lambda'] == 5)]
xian_full_ggi_4 = metrics_xian[((metrics_xian['existing_lines'] != 0) & ~np.isnan(metrics_xian['existing_lines']))
& (metrics_xian['reward'] == 'ggi')
& (metrics_xian['ggi_weight'] == 4)]
xian_full_rawls = metrics_xian[((metrics_xian['existing_lines'] != 0) & ~np.isnan(metrics_xian['existing_lines']))
& (metrics_xian['reward'] == 'rawls')]
xian_full_ses0_od, xian_full_ses0_gini, xian_full_ses0_lq = print_stats(xian_full_ses0, 'xian_full_ses_0')
xian_full_ses1_od, xian_full_ses1_gini, xian_full_ses1_lq = print_stats(xian_full_ses1, 'xian_full_ses_1')
xian_full_var_5_od, xian_full_var_5_gini, xian_full_var_5_lq = print_stats(xian_full_var_5, 'xian_full_var_5')
xian_full_rawls_od, xian_full_rawls_gini, xian_full_rawls_lq = print_stats(xian_full_rawls, 'xian_full_rawls')
xian_full_ggi_4_od, xian_full_ggi_4_gini, xian_full_ggi_4_lq = print_stats(xian_full_ggi_4, 'xian_full_ggi_4')
xian_full_plot = [
['Max. Effic.', 'xian_20220812_09_42_57.652815', cp[0], '', 'o'],
['Access. Index', 'xian_20220812_14_44_22.783845', cp[1], '-', 's'],
['Var.Reg', 'xian_20220811_22_41_02.456631', cp[2], '+', '^'],
['Rawls', 'xian_20220813_09_28_43.208981', cp[3], 'o', 'v'],
# ['GGI', 'xian_20220812_18_59_40.094535', cp[4], '/', 'D'],
['GGI', 'xian_20220814_12_10_27.594976', cp[4], '/', 'D'],
]
create_all_plots(xian, metrics_xian, xian_full_plot,
bar_plot_models=['Max. Effic.', 'Rawls', 'GGI'],
line_plot_models=['Max. Effic.', 'Access. Index', 'Var.Reg', 'Rawls', 'GGI'],
scatter_plot_models=['Max. Effic.', 'Access. Index', 'Var.Reg', 'Rawls', 'GGI'],
scatter_x=[xian_full_ses0_od, xian_full_ses1_od, xian_full_var_5_od, xian_full_rawls_od, xian_full_ggi_4_od],
scatter_y=[xian_full_ses0_gini, xian_full_ses1_gini, xian_full_var_5_gini, xian_full_rawls_gini, xian_full_ggi_4_gini],
plot_name_prefix='xian_full',
env_name="Xi'an")
#%%
# ams_empty_ggi_2[['actor_lr', 'critic_lr', 'mean_sat_group_od_pct', 'group_gini', 'budget', 'existing_lines', 'ignore_existing_lines']].sort_values('mean_sat_group_od_pct')
# #%%
# # TODO: transfer this method to the environment class.
# dilemma = Environment(Path(f"./environments/dilemma_5x5"))
# def calculate_agg_od(environment):
# """Calculate aggregate origin-destination flow matrix for each grid square of the given environment.
# Args:
# environment (Environment): environment for which to calcualte aggregate OD per grid square.
# Returns:
# torch.Tensor: aggregate od by grid
# """
# #
# # A measure of importance of each square.
# agg_od_g = torch.zeros((environment.grid_x_size, environment.grid_y_size)).to(device)
# agg_od_v = environment.od_mx.sum(axis=1)
# # Get the grid indices.
# for i in range(agg_od_v.shape[0]):
# g = environment.vector_to_grid(torch.Tensor([i])).type(torch.int32)
# agg_od_g[g[0], g[1]] = agg_od_v[i]
# return agg_od_g
# dilemma_od = calculate_agg_od(dilemma).cpu()
# fig, ax = plt.subplots(figsize=(10, 10))
# im0 = ax.imshow(dilemma_od, cm.get_cmap('Blues'))
# ax.set_xticks(np.arange(-.5, 4, 1))
# ax.set_yticks(np.arange(-.5, 4, 1))
# ax.set_xticklabels(np.arange(0, 5, 1))
# ax.set_yticklabels(np.arange(0, 5, 1))
# ax.grid(color='gray', linewidth=2)
# # cax = fig.add_axes([0.65, 0.175, 0.2, 0.02])
# fig.colorbar(im0, orientation='vertical', fraction=0.046, pad=0.04)
# ax.set_title('(A) Aggregate Origin-Destination Flow')
# # %% PLOT XIAN RESULTS
# xian = Environment(Path(f"./environments/xian/"), groups_file='price_groups_5.txt')
# # metrics_xian = prepare_metrics_df(['old_xian_16_01_47.060823', 'old_xian_16_03_51.724205', 'old_xian_10_48_03.743589', 'old_xian_16_22_50.580720'])
# metrics_xian = prepare_metrics_df(['old_xian_16_22_50.580720', 'old_xian_10_48_03.743589'])