wireworld.py

   1 #!/usr/bin/env python
   2
   3 # Any copyright is dedicated to the Public Domain.
   4 # https://creativecommons.org/publicdomain/zero/1.0/
   5
   6 # Written by Francois Fleuret <francois@fleuret.org>
   7
   8 import math, sys, tqdm, os
   9
  10 import torch, torchvision
  11
  12 from torch import nn
  13 from torch.nn import functional as F
  14
  15 ######################################################################
  16
  17 import problem
  18
  19
  20 class Wireworld(problem.Problem):
  21     colors = torch.tensor(
  22         [
  23             [128, 128, 128],
  24             [128, 128, 255],
  25             [255, 0, 0],
  26             [255, 255, 0],
  27         ]
  28     )
  29
  30     token_empty = 0
  31     token_head = 1
  32     token_tail = 2
  33     token_conductor = 3
  34     token_forward = 4
  35     token_backward = 5
  36
  37     token2char = (
  38         "_" + "".join([chr(ord("A") + n) for n in range(len(colors) - 1)]) + "><"
  39     )
  40
  41     def __init__(
  42         self, height=6, width=8, nb_objects=2, nb_walls=2, speed=1, nb_iterations=4
  43     ):
  44         self.height = height
  45         self.width = width
  46         self.nb_objects = nb_objects
  47         self.nb_walls = nb_walls
  48         self.speed = speed
  49         self.nb_iterations = nb_iterations
  50
  51     def direction_tokens(self):
  52         return self.token_forward, self.token_backward
  53
  54     def generate_frame_sequences(self, nb):
  55         frame_sequences = []
  56
  57         result = torch.full(
  58             (nb * 4, self.nb_iterations, self.height, self.width), self.token_empty
  59         )
  60
  61         for n in range(result.size(0)):
  62             while True:
  63                 i = torch.randint(self.height, (1,))
  64                 j = torch.randint(self.width, (1,))
  65                 v = torch.randint(2, (2,))
  66                 vi = v[0] * (v[1] * 2 - 1)
  67                 vj = (1 - v[0]) * (v[1] * 2 - 1)
  68                 while True:
  69                     if i < 0 or i >= self.height or j < 0 or j >= self.width:
  70                         break
  71                     result[n, 0, i, j] = self.token_conductor
  72                     i += vi
  73                     j += vj
  74                 if torch.rand(1) < 0.5:
  75                     break
  76
  77         weight = torch.full((1, 1, 3, 3), 1.0)
  78
  79         mask = (torch.rand(result[:, 0].size()) < 0.01).long()
  80         rand = torch.randint(4, mask.size())
  81         result[:, 0] = mask * rand + (1 - mask) * result[:, 0]
  82
  83         # empty->empty
  84         # head->tail
  85         # tail->conductor
  86         # conductor->head if 1 or 2 head in the neighborhood, or remains conductor
  87
  88         for l in range(self.nb_iterations * self.speed - 1):
  89             nb_head_neighbors = (
  90                 F.conv2d(
  91                     input=(result[:, l] == self.token_head).float()[:, None, :, :],
  92                     weight=weight,
  93                     padding=1,
  94                 )
  95                 .long()
  96                 .squeeze(1)
  97             )
  98             mask_1_or_2_heads = (nb_head_neighbors == 1).long() + (
  99                 nb_head_neighbors == 2
 100             ).long()
 101             result[:, l + 1] = (
 102                 (result[:, l] == self.token_empty).long() * self.token_empty
 103                 + (result[:, l] == self.token_head).long() * self.token_tail
 104                 + (result[:, l] == self.token_tail).long() * self.token_conductor
 105                 + (result[:, l] == self.token_conductor).long()
 106                 * (
 107                     mask_1_or_2_heads * self.token_head
 108                     + (1 - mask_1_or_2_heads) * self.token_conductor
 109                 )
 110             )
 111
 112         i = (result[:, -1] == self.token_head).flatten(1).max(dim=1).values > 0
 113
 114         result = result[
 115             torch.arange(self.nb_iterations, device=result.device) * self.speed
 116         ]
 117
 118         if result.size(0) < nb:
 119             # print(result.size(0))
 120             result = torch.cat(
 121                 [result, self.generate_frame_sequences(nb - result.size(0))], dim=0
 122             )
 123
 124         return result[:nb]
 125
 126     def generate_token_sequences(self, nb):
 127         frame_sequences = self.generate_frame_sequences(nb)
 128
 129         result = []
 130
 131         for frame_sequence in frame_sequences:
 132             a = []
 133             if torch.rand(1) < 0.5:
 134                 for frame in frame_sequence:
 135                     if len(a) > 0:
 136                         a.append(torch.tensor([self.token_forward]))
 137                     a.append(frame.flatten())
 138             else:
 139                 for frame in reversed(frame_sequence):
 140                     if len(a) > 0:
 141                         a.append(torch.tensor([self.token_backward]))
 142                     a.append(frame.flatten())
 143
 144             result.append(torch.cat(a, dim=0)[None, :])
 145
 146         return torch.cat(result, dim=0)
 147
 148     ######################################################################
 149
 150     def frame2img(self, x, scale=15):
 151         x = x.reshape(-1, self.height, self.width)
 152         m = torch.logical_and(x >= 0, x < 4).long()
 153
 154         x = self.colors[x * m].permute(0, 3, 1, 2)
 155         s = x.shape
 156         x = x[:, :, :, None, :, None].expand(-1, -1, -1, scale, -1, scale)
 157         x = x.reshape(s[0], s[1], s[2] * scale, s[3] * scale)
 158
 159         x[:, :, :, torch.arange(0, x.size(3), scale)] = 0
 160         x[:, :, torch.arange(0, x.size(2), scale), :] = 0
 161         x = x[:, :, 1:, 1:]
 162
 163         for n in range(m.size(0)):
 164             for i in range(m.size(1)):
 165                 for j in range(m.size(2)):
 166                     if m[n, i, j] == 0:
 167                         for k in range(2, scale - 2):
 168                             for l in [0, 1]:
 169                                 x[n, :, i * scale + k, j * scale + k - l] = 0
 170                                 x[
 171                                     n, :, i * scale + scale - 1 - k, j * scale + k - l
 172                                 ] = 0
 173
 174         return x
 175
 176     def seq2img(self, seq, scale=15):
 177         all = [
 178             self.frame2img(
 179                 seq[:, : self.height * self.width].reshape(-1, self.height, self.width),
 180                 scale,
 181             )
 182         ]
 183
 184         separator = torch.full((seq.size(0), 3, self.height * scale - 1, 1), 0)
 185
 186         t = self.height * self.width
 187
 188         while t < seq.size(1):
 189             direction_tokens = seq[:, t]
 190             t += 1
 191
 192             direction_images = self.colors[
 193                 torch.full(
 194                     (direction_tokens.size(0), self.height * scale - 1, scale), 0
 195                 )
 196             ].permute(0, 3, 1, 2)
 197
 198             for n in range(direction_tokens.size(0)):
 199                 if direction_tokens[n] == self.token_forward:
 200                     for k in range(scale):
 201                         for l in [0, 1]:
 202                             direction_images[
 203                                 n,
 204                                 :,
 205                                 (self.height * scale) // 2 - scale // 2 + k - l,
 206                                 3 + scale // 2 - abs(k - scale // 2),
 207                             ] = 0
 208                 elif direction_tokens[n] == self.token_backward:
 209                     for k in range(scale):
 210                         for l in [0, 1]:
 211                             direction_images[
 212                                 n,
 213                                 :,
 214                                 (self.height * scale) // 2 - scale // 2 + k - l,
 215                                 3 + abs(k - scale // 2),
 216                             ] = 0
 217                 else:
 218                     for k in range(2, scale - 2):
 219                         for l in [0, 1]:
 220                             direction_images[
 221                                 n,
 222                                 :,
 223                                 (self.height * scale) // 2 - scale // 2 + k - l,
 224                                 k,
 225                             ] = 0
 226                             direction_images[
 227                                 n,
 228                                 :,
 229                                 (self.height * scale) // 2 - scale // 2 + k - l,
 230                                 scale - 1 - k,
 231                             ] = 0
 232
 233             all += [
 234                 separator,
 235                 direction_images,
 236                 separator,
 237                 self.frame2img(
 238                     seq[:, t : t + self.height * self.width].reshape(
 239                         -1, self.height, self.width
 240                     ),
 241                     scale,
 242                 ),
 243             ]
 244
 245             t += self.height * self.width
 246
 247         return torch.cat(all, dim=3)
 248
 249     def seq2str(self, seq):
 250         result = []
 251         for s in seq:
 252             result.append("".join([self.token2char[v] for v in s]))
 253         return result
 254
 255     def save_image(self, input, result_dir, filename):
 256         img = self.seq2img(input.to("cpu"))
 257         image_name = os.path.join(result_dir, filename)
 258         torchvision.utils.save_image(img.float() / 255.0, image_name, nrow=6, padding=4)
 259
 260     def save_quizzes(self, input, result_dir, filename_prefix):
 261         self.save_image(input, result_dir, filename_prefix + ".png")
 262
 263
 264 ######################################################################
 265
 266 if __name__ == "__main__":
 267     import time
 268
 269     wireworld = Wireworld(height=10, width=15, nb_iterations=2, speed=1)
 270
 271     start_time = time.perf_counter()
 272     frame_sequences = wireworld.generate_frame_sequences(nb=96)
 273     delay = time.perf_counter() - start_time
 274     print(f"{frame_sequences.size(0)/delay:02f} seq/s")
 275
 276     # print(wireworld.seq2str(seq[:4]))
 277
 278     # for t in range(frame_sequences.size(1)):
 279     # img = wireworld.seq2img(frame_sequences[:, t])
 280     # torchvision.utils.save_image(
 281     # img.float() / 255.0,
 282     # f"/tmp/frame_{t:03d}.png",
 283     # nrow=8,
 284     # padding=6,
 285     # pad_value=0,
 286     # )
 287
 288     # m = (torch.rand(seq.size()) < 0.05).long()
 289     # seq = (1 - m) * seq + m * 23
 290
 291     token_sequences = wireworld.generate_token_sequences(32)
 292     wireworld.save_quizzes(token_sequences, "/tmp", "seq")
 293     # img = wireworld.seq2img(frame_sequences[:60])
 294
 295     # torchvision.utils.save_image(
 296     # img.float() / 255.0, "/tmp/world.png", nrow=6, padding=10, pad_value=0.1
 297     # )