-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathwudc.py
executable file
·331 lines (284 loc) · 12.3 KB
/
wudc.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
#!/usr/bin/python3
"""Simplified version of a Hungarian algorithm-based WUDC draw generator.
Requires the munkres package (available on PyPI).
Chuan-Zheng Lee <[email protected]>
June 2017
"""
import argparse
import munkres
import random
from collections import Counter
from itertools import chain, repeat
from math import log2
from statistics import pvariance
from badness import get_vanschelven_position_badness
munkres.DISALLOWED_PRINTVAL = "-"
DISALLOWED = munkres.DISALLOWED
m = munkres.Munkres()
def read_input_file(filename, include_all=False):
"""Reads an input file, returns a list of 3-tuples, (team, points, history).
"""
f = open(filename)
data = []
for line in f:
team, points, history, active = line.split("\t")
if not include_all and int(active) == 0:
continue
history = [int(x) for x in history.split(",")]
data.append((team, int(points), history))
if not include_all:
assert len(data) % 4 == 0, "There were %d teams" % len(data)
return data
def define_rooms(points):
"""Returns a list of tuples, each one being the team points
values that are allowed to be in that room. Example:
[(17, 16, 15), (15, 14), (15, 14), ...]
"""
assert len(points) % 4 == 0
counts = Counter(points)
rooms = []
bracket = []
nteams = 0
pullups_needed = 0
for p in range(max(points), -1, -1):
if pullups_needed < counts[p]: # complete the bracket
if pullups_needed:
bracket.append(p)
counts[p] -= pullups_needed
nteams += pullups_needed
assert nteams % 4 == 0
rooms += [tuple(bracket)] * (nteams // 4)
nteams = 0
bracket = []
# add this entire bracket to the bracket
if counts[p] > 0:
bracket.append(p)
nteams += counts[p]
pullups_needed = (-nteams) % 4
assert nteams % 4 == 0
rooms += [tuple(bracket)] * (nteams // 4)
return rooms
def profile_after(pos, profile):
new_profile = profile.copy()
new_profile[pos] += 1
return new_profile
def cost_simple(pos, profile):
return profile[pos] - min(profile)
def cost_vanschelven(pos, profile):
return get_vanschelven_position_badness(profile_after(pos, profile))
def cost_entropy(pos, profile):
profile = profile_after(pos, profile)
n = sum(profile)
probs = [p/n for p in profile]
selfinfo = [0 if p == 0 else -p*log2(p) for p in probs]
return (2 - sum(selfinfo)) * n
def cost_collision_entropy(pos, profile):
profile = profile_after(pos, profile)
n = sum(profile)
probs = [p/n for p in profile]
return (2 + log2(sum([p ** 2 for p in probs]))) * n
def cost_pvariance(pos, profile):
return pvariance(profile_after(pos, profile))
def cost_adjusted_pvariance(pos, profile):
profile = profile_after(pos, profile)
n = sum(profile)
best = [n // 4] * (4 - n % 4) + [n // 4 + 1] * (n % 4)
return pvariance(profile) - pvariance(best)
def generate_cost_matrix(data, cost_fn):
"""Returns a cost matrix for the tournament.
Rows (inner lists) are teams, in the same order as in data.
Columns (elements) are positions in rooms, ordered first by room in the
order returned by `rooms`, then in speaking order (OG, OO, CG, CO).
Rules:
- if the team (given its points) is not allowed in the room, use DISALLOWED.
- otherwise, for each position, use the position badness that would arise
if the team were allocated to that position.
"""
nteams = len(data)
assert nteams % 4 == 0
rooms = define_rooms([p for _, p, _ in data])
assert len(rooms) == nteams / 4
costs = []
for _, points, history in data:
row = []
min_hist = min(history)
for room in rooms:
if points not in room:
row.extend([DISALLOWED, DISALLOWED, DISALLOWED, DISALLOWED])
else:
row.extend([cost_fn(i, history) for i in range(4)])
assert len(row) == nteams
costs.append(row)
assert len(costs) == nteams
return costs
def hungarian_shuffled(costs):
"""Applies the Hungarian algorithm to `costs`, but permutes the rows and
columns of the matrix first. Returns a list of indices pairs (row, col)."""
n = len(costs)
I = random.sample(range(n), n)
J = random.sample(range(n), n)
C = [[costs[i][j] for j in J] for i in I]
indices = m.compute(C)
return [(I[i], J[j]) for i, j in indices]
def collate_rooms(data, indices):
rooms = [[None, None, None, None] for i in range(len(indices) // 4)]
for t, r in indices:
rooms[r // 4][r % 4] = data[t]
return rooms
def generate_draw(data, cost_fn):
costs = generate_cost_matrix(data, cost_fn)
indices = hungarian_shuffled(costs)
rooms = collate_rooms(data, indices)
return rooms
def show_rooms(rooms, color=False):
if color:
YELLOW = "\033[0;33m"
GREEN = "\033[0;32m"
NORMAL = "\033[0m"
else:
YELLOW = GREEN = NORMAL = ""
for room in rooms:
bracket = max([t[1] for t in room])
teams = []
for pos, (team, points, history) in enumerate(room):
history = history.copy()
history[pos] = GREEN + str(history[pos]) + NORMAL
points_str = str(points).rjust(2)
if points != bracket:
points_str = YELLOW + points_str + NORMAL
teams.append("{team:>12s} {points!s:>2s} {history:7s}".format(
team=team[:12], points=points_str, history=",".join(map(str, history))))
print(" ".join(teams))
print()
def compare_badness(rooms, other_filename, cost_fn, color=False, quiet=False):
"""Compares the position badness implied by `data` and `indices`, to that
stored in `other_filename`."""
other_data = read_input_file(other_filename, include_all=True)
other_histories = {team: history for team, _, history in other_data}
this_total_cost = 0
other_total_cost = 0
this_total_badness = 0
other_total_badness = 0
teams = []
for room in rooms:
for pos, (team, _, history) in enumerate(room):
this_history = history.copy()
this_history[pos] += 1
this_cost = cost_fn(pos, history)
this_badness = cost_vanschelven(pos, history)
this_total_cost += this_cost
this_total_badness += this_badness
other_pos = [x != y for x, y in zip(history, other_histories[team])].index(True)
other_cost = cost_fn(other_pos, history)
other_badness = cost_vanschelven(other_pos, history)
other_total_cost += other_cost
other_total_badness += other_badness
teams.append((team, history, this_cost, this_badness, this_history, other_cost, other_badness, other_histories[team]))
if not quiet:
teams.sort(key=lambda x: (x[3], x[6]), reverse=True)
if color:
BLUE = "\033[0;34m"
GREEN = "\033[32m"
NORMAL = "\033[0m"
BOLD_CYAN = "\033[1;36m"
BOLD_YELLOW = "\033[1;33m"
else:
BOLD_CYAN = BLUE = GREEN = BOLD_YELLOW = NORMAL = ""
def history_string(base, original, changed):
strings = []
for a, b in zip(original, changed):
if a != b:
strings.append(GREEN + str(b) + base)
else:
strings.append(str(b))
return base + ",".join(strings) + NORMAL
print(BOLD_CYAN + " team ours original" + NORMAL)
for team, original_history, this_cost, this_badness, this_history, other_cost, other_badness, other_history in teams:
this_base = BLUE if this_badness == 0 else BOLD_YELLOW if this_badness > other_badness else NORMAL
other_base = BLUE if other_badness == 0 else BOLD_YELLOW if other_badness > this_badness else NORMAL
cost_format = "2d" if isinstance(this_cost, int) else "4.2f"
this_cost_str = ("{c}{cost:>" + cost_format + "}{n}").format(cost=this_cost, c=this_base, n=NORMAL)
other_cost_str = ("{c}{cost:>" + cost_format + "}{n}").format(cost=other_cost, c=other_base, n=NORMAL)
this_badness_str = "{c}({bad:>2d}){n}".format(bad=this_badness, c=this_base, n=NORMAL)
other_badness_str = "{c}({bad:>2d}){n}".format(bad=other_badness, c=other_base, n=NORMAL)
this_history_str = history_string(this_base, original_history, this_history)
other_history_str = history_string(other_base, original_history, other_history)
print("{team:>17s}: {cost1:>2s} {bad1:>2s} {hist1:7s} {cost2:>2s} {bad2:>2s} {hist2:7s}".format(
team=team[:17], bad1=this_badness_str, bad2=other_badness_str,
cost1=this_cost_str, cost2=other_cost_str,
hist1=this_history_str, hist2=other_history_str,
))
return this_total_cost, other_total_cost, this_total_badness, other_total_badness
def show_original_rooms(data, filename, color=False):
properties = {team: (points, history) for team, points, history in data}
f = open(filename)
rooms = []
for line in f:
names = line.split("\t")
rooms.append([(name.strip(),) + properties[name.strip()] for name in names])
rooms.sort(key=lambda x: max(y[1] for y in x), reverse=True)
show_rooms(rooms, color)
def _print_heading(message, color=False):
if color:
print("\033[1;36m" + message + "\033[0m")
else:
print(message)
COST_FUNCTIONS = {
"simple": cost_simple,
"vanschelven": cost_vanschelven,
"entropy": cost_entropy,
"collision": cost_collision_entropy,
"pvar": cost_pvariance,
"adjpvar": cost_adjusted_pvariance,
}
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("tournament")
parser.add_argument("round", type=int, nargs='?', default=None)
parser.add_argument("-q", "--quiet", help="Print only the final costs", action="store_true", default=False)
parser.add_argument("-C", "--compare-file")
parser.add_argument("-D", "--actual-draw")
parser.add_argument("-m", "--no-color", dest="color", action="store_false", default=True)
parser.add_argument("-c", "--cost-method", choices=COST_FUNCTIONS.keys(), default="vanschelven")
parser.add_argument("-e", "--exponent", type=float, default=None,
help=("If specified, the cost function is raised to this exponent."))
args = parser.parse_args()
import os.path
if os.path.isfile(args.tournament):
filename = args.tournament
comparefile = args.compare_file
actualdrawfile = args.actual_draw
elif args.round is not None:
filename = os.path.join("data", args.tournament, "round{:d}_after.tsv".format(args.round - 1))
comparefile = os.path.join("data", args.tournament, "round{:d}_after.tsv".format(args.round))
actualdrawfile = os.path.join("data", args.tournament, "round{:d}_actual.tsv".format(args.round))
else:
print("Either the first argument must be a file name, or the second argument must be a round number.")
exit()
if args.exponent:
exp = float(args.exponent)
def cost_fn(pos, profile):
return COST_FUNCTIONS[args.cost_method](pos, profile) ** exp
else:
cost_fn = COST_FUNCTIONS[args.cost_method]
data = read_input_file(filename)
rooms = generate_draw(data, cost_fn)
if not args.quiet:
_print_heading("Our draw:", args.color)
show_rooms(rooms, args.color)
if actualdrawfile:
_print_heading("\033[1;36mOriginal draw:\033[0m", args.color)
show_original_rooms(data, actualdrawfile, args.color)
if comparefile:
this_cost, other_cost, this_badness, other_badness = compare_badness(rooms, comparefile, cost_fn, args.color, args.quiet)
if args.color:
CYAN = "\033[1;36m"
BOLD_WHITE = "\033[1;37m"
NORMAL = "\033[0m"
else:
CYAN = BOLD_WHITE = NORMAL = ""
print()
print(CYAN + " our total cost:" + BOLD_WHITE, this_cost, NORMAL)
print(CYAN + " original total cost:" + BOLD_WHITE, other_cost, NORMAL)
print(CYAN + " our total badness:" + BOLD_WHITE, this_badness, NORMAL)
print(CYAN + "original total badness:" + BOLD_WHITE, other_badness, NORMAL)