File size: 36,298 Bytes
f69e256 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 | """Traverse Ordering (two-curve variant).
Two smooth tangled curves cross each other multiple times (perimeter-anchored
turtle walks, same generator style as line_intersections). Labeled marks
(A, B, C, ...) are placed on BOTH curves at well-spaced arc-length positions,
avoiding crossings and close-strand regions. Exactly one curve additionally
has a green 'S' marker at one of its endpoints. The task: starting from S,
follow that curve to its other endpoint and list the labels of marks on the
S-curve in the order visited. Marks on the other curve are distractors.
"""
from __future__ import annotations
import argparse
import json
import math
import os
import random
import string
import sys
from collections import defaultdict
from pathlib import Path
from typing import Dict, List, Tuple
import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt
import numpy as np
from tqdm import tqdm
LINE_COLOR = "#2f2f2f"
MARK_COLORS = [
"#e63946", "#457b9d", "#2a9d8f", "#e9c46a", "#f4a261",
"#264653", "#6a4c93", "#1982c4", "#8ac926", "#ff595e",
]
S_COLOR = "#1f9d3f"
# ── Perimeter-anchored turtle curve (from line_intersections) ──────
def _perimeter_point(s: float, width: int, height: int, margin: int
) -> Tuple[np.ndarray, np.ndarray]:
side = int(s) % 4
t = s - int(s)
if side == 0:
x = margin + (width - 2 * margin) * t
y = margin
inward = np.array([0.0, 1.0])
elif side == 1:
x = width - margin
y = margin + (height - 2 * margin) * t
inward = np.array([-1.0, 0.0])
elif side == 2:
x = width - margin - (width - 2 * margin) * t
y = height - margin
inward = np.array([0.0, -1.0])
else:
x = margin
y = height - margin - (height - 2 * margin) * t
inward = np.array([1.0, 0.0])
return np.array([x, y]), inward
def sample_perimeter_anchors(
rng: random.Random,
num_anchors: int,
width: int,
height: int,
margin: int,
min_gap: float | None = None,
jitter_deg: float = 10.0,
max_attempts: int = 300,
) -> List[Tuple[np.ndarray, np.ndarray]]:
if min_gap is None:
min_gap = (4.0 / num_anchors) * 0.5
for _ in range(max_attempts):
positions = sorted(rng.uniform(0.0, 4.0) for _ in range(num_anchors))
ok = True
for i in range(num_anchors):
gap = (positions[(i + 1) % num_anchors] - positions[i]) % 4.0
if gap < min_gap:
ok = False
break
if ok:
anchors = []
for s in positions:
pt, inward = _perimeter_point(s, width, height, margin)
ang = math.atan2(inward[1], inward[0])
ang += math.radians(rng.uniform(-jitter_deg, jitter_deg))
inward = np.array([math.cos(ang), math.sin(ang)])
anchors.append((pt, inward))
return anchors
return None
def _angle_diff(current: float, target: float) -> float:
d = target - current
return (d + math.pi) % (2 * math.pi) - math.pi
def build_anchored_curve(
rng: random.Random,
width: int,
height: int,
start_pos: np.ndarray,
start_inward: np.ndarray,
end_pos: np.ndarray,
end_inward: np.ndarray,
leader_length: float = 150.0,
step_size: float = 3.3,
max_turn: float = 0.028,
drift_rate: float = 0.007,
num_waypoints: int = 2,
interior_margin: int = 260,
max_steps: int = 1800,
) -> np.ndarray:
pts: List[np.ndarray] = [start_pos.copy()]
x, y = float(start_pos[0]), float(start_pos[1])
theta = math.atan2(start_inward[1], start_inward[0])
raw_wps: List[np.ndarray] = []
for _ in range(num_waypoints):
wx = rng.uniform(interior_margin, width - interior_margin)
wy = rng.uniform(interior_margin, height - interior_margin)
raw_wps.append(np.array([wx, wy]))
direction = end_pos - start_pos
dir_norm = direction / (np.linalg.norm(direction) + 1e-9)
raw_wps.sort(key=lambda p: float(np.dot(p - start_pos, dir_norm)))
funnel = end_pos + end_inward * 400.0
lookahead = end_pos + (-end_inward) * 240.0
waypoints: List[np.ndarray] = raw_wps + [funnel, lookahead]
exit_theta = math.atan2(-end_inward[1], -end_inward[0])
entry_steps = max(2, int(leader_length / step_size))
wp_idx = 0
drift = 0.0
perp = np.array([-end_inward[1], end_inward[0]])
for step_count in range(max_steps):
sd = ((x - end_pos[0]) * end_inward[0]
+ (y - end_pos[1]) * end_inward[1])
lat = abs((x - end_pos[0]) * perp[0] + (y - end_pos[1]) * perp[1])
if sd <= 0.0:
break
if step_count < entry_steps:
turn = 0.0
elif sd < leader_length and lat < 40.0:
head_err = _angle_diff(theta, exit_theta)
turn = max(-max_turn, min(max_turn, head_err))
else:
is_final = (wp_idx == len(waypoints) - 1)
if is_final:
wx, wy = waypoints[-1]
pull = 0.18
else:
wx, wy = waypoints[wp_idx]
steer_dist = math.hypot(wx - x, wy - y)
if steer_dist < 140.0:
wp_idx += 1
continue
pull = 0.09
drift += rng.gauss(0, drift_rate)
drift = max(-max_turn * 0.7, min(max_turn * 0.7, drift))
turn = drift + rng.gauss(0, max_turn * 0.12)
target_ang = math.atan2(wy - y, wx - x)
steer = _angle_diff(theta, target_ang)
turn += steer * pull
turn = max(-max_turn, min(max_turn, turn))
theta += turn
x += step_size * math.cos(theta)
y += step_size * math.sin(theta)
pts.append(np.array([x, y]))
return np.array(pts)
# ── Intersection detection ─────────────────────────────────────────
def _segments_cross(p0, p1, q0, q1) -> bool:
eps = 1e-8
o1 = (p1[0]-p0[0])*(q0[1]-p0[1]) - (p1[1]-p0[1])*(q0[0]-p0[0])
o2 = (p1[0]-p0[0])*(q1[1]-p0[1]) - (p1[1]-p0[1])*(q1[0]-p0[0])
o3 = (q1[0]-q0[0])*(p0[1]-q0[1]) - (q1[1]-q0[1])*(p0[0]-q0[0])
o4 = (q1[0]-q0[0])*(p1[1]-q0[1]) - (q1[1]-q0[1])*(p1[0]-q0[0])
return ((o1 > eps and o2 < -eps) or (o1 < -eps and o2 > eps)) and \
((o3 > eps and o4 < -eps) or (o3 < -eps and o4 > eps))
def _crossing_angle(p0, p1, q0, q1) -> float:
d1 = p1 - p0
d2 = q1 - q0
cos_a = np.dot(d1, d2) / (np.linalg.norm(d1) * np.linalg.norm(d2) + 1e-12)
return math.degrees(math.acos(min(1.0, abs(cos_a))))
def _find_crossings(
poly_a: np.ndarray,
poly_b: np.ndarray,
same_curve: bool = False,
min_seg_gap: int = 10,
):
na = len(poly_a) - 1
nb = len(poly_b) - 1
a_min_x = np.minimum(poly_a[:-1, 0], poly_a[1:, 0])
a_max_x = np.maximum(poly_a[:-1, 0], poly_a[1:, 0])
a_min_y = np.minimum(poly_a[:-1, 1], poly_a[1:, 1])
a_max_y = np.maximum(poly_a[:-1, 1], poly_a[1:, 1])
b_min_x = np.minimum(poly_b[:-1, 0], poly_b[1:, 0])
b_max_x = np.maximum(poly_b[:-1, 0], poly_b[1:, 0])
b_min_y = np.minimum(poly_b[:-1, 1], poly_b[1:, 1])
b_max_y = np.maximum(poly_b[:-1, 1], poly_b[1:, 1])
all_min_x = np.concatenate([a_min_x, b_min_x])
all_max_x = np.concatenate([a_max_x, b_max_x])
cell_size = max(np.median(all_max_x - all_min_x), 1.0) * 3
grid_a = defaultdict(list)
for i in range(na):
cx0 = int(a_min_x[i] / cell_size); cx1 = int(a_max_x[i] / cell_size)
cy0 = int(a_min_y[i] / cell_size); cy1 = int(a_max_y[i] / cell_size)
for gx in range(cx0, cx1 + 1):
for gy in range(cy0, cy1 + 1):
grid_a[(gx, gy)].append(i)
grid_b = defaultdict(list)
for j in range(nb):
cx0 = int(b_min_x[j] / cell_size); cx1 = int(b_max_x[j] / cell_size)
cy0 = int(b_min_y[j] / cell_size); cy1 = int(b_max_y[j] / cell_size)
for gx in range(cx0, cx1 + 1):
for gy in range(cy0, cy1 + 1):
grid_b[(gx, gy)].append(j)
checked = set()
worst_angle = 180.0
details: List[Dict] = []
for cell_key in grid_a:
if cell_key not in grid_b:
continue
for i in grid_a[cell_key]:
for j in grid_b[cell_key]:
if same_curve:
ii, jj = min(i, j), max(i, j)
if jj - ii < min_seg_gap:
continue
key = (ii, jj)
else:
key = (i, j)
if key in checked:
continue
checked.add(key)
if a_max_x[i] < b_min_x[j] or b_max_x[j] < a_min_x[i] or \
a_max_y[i] < b_min_y[j] or b_max_y[j] < a_min_y[i]:
continue
if same_curve:
si, sj = key
p0, p1 = poly_a[si], poly_a[si + 1]
q0, q1 = poly_b[sj], poly_b[sj + 1]
else:
si, sj = i, j
p0, p1 = poly_a[si], poly_a[si + 1]
q0, q1 = poly_b[sj], poly_b[sj + 1]
if not _segments_cross(p0, p1, q0, q1):
continue
d1 = p1 - p0
d2 = q1 - q0
denom = d1[0] * d2[1] - d1[1] * d2[0]
if abs(denom) < 1e-12:
continue
ti = ((q0[0] - p0[0]) * d2[1] - (q0[1] - p0[1]) * d2[0]) / denom
tj = ((q0[0] - p0[0]) * d1[1] - (q0[1] - p0[1]) * d1[0]) / denom
px = float(p0[0] + ti * d1[0])
py = float(p0[1] + ti * d1[1])
angle = _crossing_angle(p0, p1, q0, q1)
worst_angle = min(worst_angle, angle)
details.append({
"i": int(si), "j": int(sj),
"ti": float(ti), "tj": float(tj),
"px": px, "py": py, "angle": float(angle),
})
return len(details), worst_angle, details
def _details_to_points(details: List[Dict]) -> np.ndarray:
if not details:
return np.zeros((0, 2))
return np.array([[d["px"], d["py"]] for d in details])
def _curves_too_close(
poly_a: np.ndarray,
poly_b: np.ndarray,
same_curve: bool = False,
min_dist: float = 7.0,
sample_step: int = 3,
self_index_gap: int = 50,
known_crossings: np.ndarray | None = None,
crossing_exclude_radius: float = 30.0,
) -> bool:
nb = len(poly_b) - 1
b_pts = poly_b[:-1]
b_vecs = poly_b[1:] - poly_b[:-1]
b_lens_sq = np.maximum((b_vecs ** 2).sum(axis=1), 1e-12)
has_crossings = known_crossings is not None and len(known_crossings) > 0
for idx in range(0, len(poly_a), sample_step):
px, py = poly_a[idx]
p = np.array([px, py])
dp = p - b_pts
t = (dp * b_vecs).sum(axis=1) / b_lens_sq
t = np.clip(t, 0.0, 1.0)
proj = b_pts + t[:, None] * b_vecs
dists = np.sqrt(((p - proj) ** 2).sum(axis=1))
if same_curve:
mask = np.abs(np.arange(nb) - idx) < self_index_gap
dists[mask] = 9999.0
if dists.min() < min_dist:
if has_crossings:
cross_dists = np.sqrt(((p - known_crossings) ** 2).sum(axis=1))
if cross_dists.min() < crossing_exclude_radius:
continue
return True
return False
# ── Mark placement ─────────────────────────────────────────────────
def place_marks_on_curve(
rng: random.Random,
polyline: np.ndarray,
other_polyline: np.ndarray,
num_marks: int,
avoid_points: List[np.ndarray],
existing_marks: List[Tuple[np.ndarray, int]] | None = None,
min_arc_frac: float = 0.10,
min_pixel_to_crossing: float = 70.0,
min_pixel_to_other_strand: float = 36.0,
min_pixel_between_marks: float = 90.0,
endpoint_arc_margin_frac: float = 0.08,
max_iter: int = 1200,
extra_polylines: List[np.ndarray] | None = None,
) -> List[int]:
"""Place `num_marks` indices along `polyline`, avoiding crossing points,
regions where the OTHER curve passes close, and existing marks on either
curve. Endpoints are reserved (for the S marker).
"""
n = len(polyline)
diffs = np.diff(polyline, axis=0)
seg_lens = np.sqrt((diffs ** 2).sum(axis=1))
arc = np.concatenate([[0.0], np.cumsum(seg_lens)])
total_len = arc[-1]
min_arc_dist = total_len * min_arc_frac
margin_arc = total_len * endpoint_arc_margin_frac
# Precompute other curve's segments for fast distance queries.
op = other_polyline
o_pts = op[:-1]
o_vecs = op[1:] - op[:-1]
o_lens_sq = np.maximum((o_vecs ** 2).sum(axis=1), 1e-12)
def dist_to_other(pt: np.ndarray) -> float:
dp = pt - o_pts
t = (dp * o_vecs).sum(axis=1) / o_lens_sq
t = np.clip(t, 0.0, 1.0)
proj = o_pts + t[:, None] * o_vecs
return float(np.sqrt(((pt - proj) ** 2).sum(axis=1)).min())
extra_bundles = []
for ep in (extra_polylines or []):
e_pts = ep[:-1]
e_vecs = ep[1:] - ep[:-1]
e_lens_sq = np.maximum((e_vecs ** 2).sum(axis=1), 1e-12)
extra_bundles.append((e_pts, e_vecs, e_lens_sq))
def dist_to_any_extra(pt: np.ndarray) -> float:
if not extra_bundles:
return 1e9
best = 1e9
for e_pts, e_vecs, e_lens_sq in extra_bundles:
dp = pt - e_pts
t = (dp * e_vecs).sum(axis=1) / e_lens_sq
t = np.clip(t, 0.0, 1.0)
proj = e_pts + t[:, None] * e_vecs
d = float(np.sqrt(((pt - proj) ** 2).sum(axis=1)).min())
if d < best:
best = d
return best
def too_near_avoid(pt: np.ndarray) -> bool:
for cp in avoid_points:
if np.sqrt(((pt - cp) ** 2).sum()) < min_pixel_to_crossing:
return True
return False
existing = existing_marks or []
mark_indices: List[int] = []
for _ in range(max_iter):
if len(mark_indices) == num_marks:
break
t = rng.uniform(margin_arc, total_len - margin_arc)
idx = int(np.searchsorted(arc, t))
idx = max(1, min(n - 2, idx))
pt = polyline[idx]
if too_near_avoid(pt):
continue
if dist_to_other(pt) < min_pixel_to_other_strand:
continue
if dist_to_any_extra(pt) < min_pixel_to_other_strand:
continue
if any(abs(arc[idx] - arc[m]) < min_arc_dist for m in mark_indices):
continue
if any(np.sqrt(((pt - polyline[m]) ** 2).sum()) < min_pixel_between_marks
for m in mark_indices):
continue
if any(np.sqrt(((pt - ep) ** 2).sum()) < min_pixel_between_marks
for ep, _ in existing):
continue
mark_indices.append(idx)
mark_indices.sort()
return mark_indices
def label_anchor_for_mark(
polyline: np.ndarray,
other_polyline: np.ndarray,
mark_idx: int,
width: int,
height: int,
offset: float = 48.0,
tangent_span: int = 6,
) -> Tuple[float, float]:
n = len(polyline)
lo = max(0, mark_idx - tangent_span)
hi = min(n - 1, mark_idx + tangent_span)
tan = polyline[hi] - polyline[lo]
tnorm = math.hypot(float(tan[0]), float(tan[1]))
if tnorm < 1e-8:
ux, uy = 0.0, -1.0
else:
ux = float(tan[0]) / tnorm
uy = float(tan[1]) / tnorm
perps = [(-uy, ux), (uy, -ux)]
pt = polyline[mark_idx]
best = None
best_score = -1e9
for px, py in perps:
cx = float(pt[0]) + px * offset
cy = float(pt[1]) + py * offset
score = 0.0
d_self = float(np.sqrt(((polyline - np.array([cx, cy])) ** 2).sum(axis=1)).min())
d_other = float(np.sqrt(((other_polyline - np.array([cx, cy])) ** 2).sum(axis=1)).min())
score += min(d_self, d_other)
if 30 < cx < width - 30 and 30 < cy < height - 30:
score += 200.0
if score > best_score:
best_score = score
best = (cx, cy)
cx, cy = best
return (
float(np.clip(cx, 34, width - 34)),
float(np.clip(cy, 34, height - 34)),
)
# ── Instance sampling ──────────────────────────────────────────────
def sample_instance(
rng: random.Random,
width: int,
height: int,
min_marks_total: int = 4,
max_marks_total: int = 7,
min_crossings: int = 7,
max_crossings: int = 25,
min_pixel_between_marks: float = 90.0,
s_endpoint_excl_px: float = 90.0,
perimeter_margin: int = 70,
interior_margin: int = 260,
max_attempts: int = 2500,
num_waypoints: int = 5,
s_marks_override: int | None = None,
o_marks_override: int | None = None,
num_extra_distractor_curves: int = 0,
extra_distractor_max_attempts: int = 80,
) -> Dict | None:
labels_pool = list(string.ascii_uppercase)
# reserve S
labels_pool = [l for l in labels_pool if l != "S"]
for _ in range(max_attempts):
anchors = sample_perimeter_anchors(
rng, 4, width, height, margin=perimeter_margin,
)
if anchors is None:
continue
# Pair anchor i with anchor i+2 (opposite-side pairing).
pairs = [(anchors[0], anchors[2]), (anchors[1], anchors[3])]
rng.shuffle(pairs)
polylines = []
ok = True
for (sp, sd), (ep, ed) in pairs:
poly = build_anchored_curve(
rng, width, height,
start_pos=sp, start_inward=sd,
end_pos=ep, end_inward=ed,
leader_length=150.0,
step_size=rng.uniform(3.0, 3.8),
max_turn=rng.uniform(0.034, 0.044),
drift_rate=rng.uniform(0.010, 0.016),
num_waypoints=num_waypoints,
interior_margin=interior_margin,
max_steps=3200,
)
if len(poly) < 50:
ok = False
break
polylines.append(poly)
if not ok:
continue
# Crossings between the two curves (no self-crossings desired, but allow).
pair_count, pair_min_angle, pair_details = _find_crossings(
polylines[0], polylines[1]
)
self_counts = []
self_details_list = []
self_min_angle = 180.0
for p in polylines:
sc, sa, sd_ = _find_crossings(p, p, same_curve=True)
self_counts.append(sc)
self_details_list.append(sd_)
self_min_angle = min(self_min_angle, sa)
total_cross = pair_count + sum(self_counts)
if pair_count < min_crossings or pair_count > max_crossings:
continue
# Keep self-crossings modest so traversal order stays clear.
if sum(self_counts) > 3:
continue
min_angle = min(pair_min_angle, self_min_angle)
if min_angle < 30.0:
continue
# Hygiene: no close strands outside crossing regions.
all_cross_pts = _details_to_points(pair_details)
# 1% of canvas short side keeps strands visibly separated at any
# resolution while staying tractable for the sampler.
min_dist_px = 0.01 * float(min(width, height))
if _curves_too_close(polylines[0], polylines[1],
min_dist=min_dist_px,
known_crossings=all_cross_pts):
continue
too_close_self = False
for i, p in enumerate(polylines):
if _curves_too_close(p, p, same_curve=True,
min_dist=min_dist_px,
known_crossings=_details_to_points(self_details_list[i])):
too_close_self = True
break
if too_close_self:
continue
# Extra unlabeled distractor curves (difficulty-driven).
extra_polylines: List[np.ndarray] = []
extra_pair_details_all: List[Dict] = []
extra_self_details_all: List[Dict] = []
extra_ok = True
for _extra_idx in range(num_extra_distractor_curves):
placed = False
for _attempt in range(extra_distractor_max_attempts):
extra_anchors = sample_perimeter_anchors(
rng, 2, width, height, margin=perimeter_margin,
)
if extra_anchors is None:
continue
(esp, esd), (eep, eed) = extra_anchors[0], extra_anchors[1]
epoly = build_anchored_curve(
rng, width, height,
start_pos=esp, start_inward=esd,
end_pos=eep, end_inward=eed,
leader_length=150.0,
step_size=rng.uniform(3.0, 3.8),
max_turn=rng.uniform(0.034, 0.044),
drift_rate=rng.uniform(0.010, 0.016),
num_waypoints=num_waypoints,
interior_margin=interior_margin,
max_steps=3200,
)
if len(epoly) < 50:
continue
# Self-crossing hygiene for the extra curve.
esc, esa, esd_det = _find_crossings(epoly, epoly, same_curve=True)
if esc > 3 or esa < 30.0:
continue
if _curves_too_close(epoly, epoly, same_curve=True,
min_dist=min_dist_px,
known_crossings=_details_to_points(esd_det)):
continue
# Cross-curve hygiene vs all existing polylines.
all_existing = polylines + extra_polylines
bad = False
candidate_pair_details: List[Dict] = []
for existing_poly in all_existing:
ec, ea, ed_det = _find_crossings(epoly, existing_poly)
if ea < 30.0:
bad = True
break
if _curves_too_close(epoly, existing_poly,
min_dist=min_dist_px,
known_crossings=_details_to_points(ed_det)):
bad = True
break
candidate_pair_details.extend(ed_det)
if bad:
continue
extra_polylines.append(epoly)
extra_pair_details_all.extend(candidate_pair_details)
extra_self_details_all.extend(esd_det)
placed = True
break
if not placed:
extra_ok = False
break
if not extra_ok:
continue
# Collect points to avoid for mark placement: all crossings.
avoid_pts: List[np.ndarray] = [np.array([d["px"], d["py"]])
for d in pair_details]
for sd_ in self_details_list:
avoid_pts.extend(np.array([d["px"], d["py"]]) for d in sd_)
for d_ in extra_pair_details_all:
avoid_pts.append(np.array([d_["px"], d_["py"]]))
for d_ in extra_self_details_all:
avoid_pts.append(np.array([d_["px"], d_["py"]]))
# Pick S-curve and which endpoint is S.
s_curve = rng.randint(0, 1)
s_at_start = rng.random() < 0.5
other_curve = 1 - s_curve
# Decide mark counts. Default: 10 total, 5 on S-curve, 5 on the
# distractor curve. If the caller narrows the range we split roughly
# evenly, keeping at least 2 on each curve.
total_marks = rng.randint(min_marks_total, max_marks_total)
s_marks = total_marks // 2
s_marks = max(2, min(s_marks, total_marks - 2))
o_marks = total_marks - s_marks
s_poly = polylines[s_curve]
o_poly = polylines[other_curve]
s_indices = place_marks_on_curve(
rng, s_poly, o_poly, s_marks, avoid_points=avoid_pts,
min_pixel_between_marks=min_pixel_between_marks,
extra_polylines=extra_polylines,
)
if len(s_indices) < s_marks:
continue
# Existing marks (pixel-space) to avoid from other curve's perspective.
existing_for_other = [(s_poly[i], i) for i in s_indices]
o_indices = place_marks_on_curve(
rng, o_poly, s_poly, o_marks, avoid_points=avoid_pts,
existing_marks=existing_for_other,
min_pixel_between_marks=min_pixel_between_marks,
extra_polylines=extra_polylines,
)
if len(o_indices) < o_marks:
continue
# Also avoid marks too close to the S endpoint on s_poly.
s_endpoint_pt = s_poly[0] if s_at_start else s_poly[-1]
all_mark_pts = [s_poly[i] for i in s_indices] + [o_poly[i] for i in o_indices]
if any(np.sqrt(((s_endpoint_pt - p) ** 2).sum()) < s_endpoint_excl_px
for p in all_mark_pts):
continue
# Assign labels. Pool of letters sized to total_marks, shuffled.
pool = labels_pool[:total_marks]
rng.shuffle(pool)
s_labels = pool[:s_marks]
o_labels = pool[s_marks:s_marks + o_marks]
# Determine traversal order on the S-curve.
# s_indices is ascending in polyline order. If s_at_start, forward;
# else reverse.
order_pairs = list(zip(s_indices, s_labels))
if not s_at_start:
order_pairs = order_pairs[::-1]
answer_sequence = [lab for _, lab in order_pairs]
answer = ", ".join(answer_sequence)
# Build mark info.
marks_info: List[Dict] = []
for idx, lab in zip(s_indices, s_labels):
lx, ly = label_anchor_for_mark(s_poly, o_poly, idx, width, height)
marks_info.append({
"label": lab,
"curve": s_curve,
"on_s_curve": True,
"polyline_index": int(idx),
"x": round(float(s_poly[idx, 0]), 2),
"y": round(float(s_poly[idx, 1]), 2),
"label_x": round(lx, 2),
"label_y": round(ly, 2),
})
for idx, lab in zip(o_indices, o_labels):
lx, ly = label_anchor_for_mark(o_poly, s_poly, idx, width, height)
marks_info.append({
"label": lab,
"curve": other_curve,
"on_s_curve": False,
"polyline_index": int(idx),
"x": round(float(o_poly[idx, 0]), 2),
"y": round(float(o_poly[idx, 1]), 2),
"label_x": round(lx, 2),
"label_y": round(ly, 2),
})
# S marker info.
s_pt = s_poly[0] if s_at_start else s_poly[-1]
# Label offset for S: away from curve tangent.
s_tan_idx = 10 if s_at_start else len(s_poly) - 11
s_idx_for_anchor = 0 if s_at_start else len(s_poly) - 1
slx, sly = label_anchor_for_mark(s_poly, o_poly, s_idx_for_anchor,
width, height, offset=52.0,
tangent_span=10)
all_labels_display = sorted(s_labels + o_labels)
question = (
"The image shows two smooth curves that cross each other. "
f"Labeled points ({', '.join(all_labels_display)}) are marked on "
"both curves. A single green 'S' marks the start endpoint of one "
"of the curves. Starting from S, follow THAT curve (the one S "
"lies on) all the way to its other endpoint, and list the labels "
"of the marked points you visit in order, separated by commas "
"(for example: B, D, A). Ignore any labeled points that lie on "
"the OTHER curve. Provide your final answer enclosed in "
"<answer>...</answer> tags."
)
all_polylines = polylines + extra_polylines
return {
"width": width,
"height": height,
"polylines": all_polylines,
"num_extra_distractor_curves": int(len(extra_polylines)),
"s_curve": int(s_curve),
"s_at_start": bool(s_at_start),
"s_point": {
"x": round(float(s_pt[0]), 2),
"y": round(float(s_pt[1]), 2),
"label_x": round(float(slx), 2),
"label_y": round(float(sly), 2),
},
"marks": marks_info,
"num_marks_total": total_marks,
"num_marks_on_s": s_marks,
"num_pair_crossings": int(pair_count),
"num_self_crossings": int(sum(self_counts)),
"question": question,
"answer": answer,
}
return None
# ── Rendering ──────────────────────────────────────────────────────
def render_instance(out_path: Path, record: Dict, noise_seed: int) -> None:
width = int(record["width"])
height = int(record["height"])
polylines = record["polylines"]
marks = record["marks"]
s_point = record["s_point"]
fig = plt.figure(figsize=(width / 100, height / 100), dpi=100)
ax = fig.add_axes([0, 0, 1, 1])
ax.set_xlim(0, width)
ax.set_ylim(height, 0)
ax.axis("off")
ax.set_facecolor("#f3efe8")
nrng = np.random.default_rng(noise_seed)
noise = nrng.normal(0.0, 1.0, size=(height, width))
noise = (noise - noise.min()) / max(noise.max() - noise.min(), 1e-6)
ax.imshow(noise, cmap="Greys", alpha=0.06, extent=(0, width, height, 0),
interpolation="bilinear")
for i, poly in enumerate(polylines):
ax.plot(poly[:, 0], poly[:, 1],
color=LINE_COLOR, linewidth=2.8,
solid_capstyle="round", solid_joinstyle="round",
zorder=2.0 + i * 0.05)
# Subtle dots at all four endpoints.
for poly in polylines:
for ep in (poly[0], poly[-1]):
ax.scatter([ep[0]], [ep[1]], s=30,
facecolors="#f3efe8", edgecolors="#999",
linewidths=1.0, zorder=3.2)
# Marks with offset labels.
for i, mark in enumerate(marks):
x, y = mark["x"], mark["y"]
lx = mark.get("label_x", x)
ly = mark.get("label_y", y - 22)
color = MARK_COLORS[i % len(MARK_COLORS)]
ax.plot([x, lx], [y, ly],
color="#9b7b56", linewidth=1.0, alpha=0.7,
linestyle=(0, (2.0, 3.2)), zorder=3.8)
ax.scatter([x], [y], s=120, facecolors=color, edgecolors="white",
linewidths=1.5, zorder=4.0)
ax.text(lx, ly, mark["label"],
fontsize=18, fontweight="bold", color=color,
ha="center", va="center", zorder=5,
bbox=dict(facecolor="#f3efe8", edgecolor=color,
boxstyle="round,pad=0.25", alpha=0.95, linewidth=1.2))
# S marker on the S-curve endpoint — green, larger.
sx, sy = s_point["x"], s_point["y"]
slx, sly = s_point["label_x"], s_point["label_y"]
ax.plot([sx, slx], [sy, sly],
color=S_COLOR, linewidth=1.2, alpha=0.8,
linestyle=(0, (2.0, 3.2)), zorder=4.2)
ax.scatter([sx], [sy], s=180, facecolors=S_COLOR, edgecolors="white",
linewidths=2.0, zorder=5.0)
ax.text(slx, sly, "S",
fontsize=20, fontweight="bold", color=S_COLOR,
ha="center", va="center", zorder=6,
bbox=dict(facecolor="#f3efe8", edgecolor=S_COLOR,
boxstyle="round,pad=0.3", alpha=0.95, linewidth=1.5))
fig.savefig(out_path, dpi=100, bbox_inches="tight", pad_inches=0)
plt.close(fig)
# ── Main ───────────────────────────────────────────────────────────
def _polyline_to_list(p: np.ndarray) -> List[List[float]]:
return [[round(float(x), 2), round(float(y), 2)] for x, y in p]
def main():
parser = argparse.ArgumentParser()
parser.add_argument("--output-root", type=Path, required=True)
parser.add_argument("--count", type=int, default=6)
parser.add_argument("--seed", type=int, default=42)
parser.add_argument("--width", type=int, default=1024)
parser.add_argument("--height", type=int, default=1024)
parser.add_argument("--min-marks", type=int, default=10)
parser.add_argument("--max-marks", type=int, default=10)
parser.add_argument("--min-crossings", type=int, default=7)
parser.add_argument("--max-crossings", type=int, default=25)
parser.add_argument("--difficulty", type=int, default=5,
help="Integer difficulty >=0; scales crossings and marks.")
parser.add_argument("--workers", type=int, default=8)
args = parser.parse_args()
d = max(0, int(args.difficulty))
# Defaults (used whether difficulty override fires or not).
min_pixel_between_marks = 90.0
s_endpoint_excl_px = 90.0
num_waypoints = 5
num_extra_distractor_curves = 0
if d > 0:
args.min_crossings = 4 + d
args.max_crossings = 6 + 2 * d
s_marks = 4 + d // 2
o_marks = 4 + d // 2
args.min_marks = s_marks + o_marks
args.max_marks = s_marks + o_marks
num_waypoints = 5 + d
min_pixel_between_marks = float(max(55, 90 - 5 * d))
s_endpoint_excl_px = float(max(50, 90 - 5 * d))
num_extra_distractor_curves = d // 3
# Canvas scaling with difficulty (sqrt of mark-density growth).
N_d = 8 + 4 * d
N_0 = 8
s_scale = math.sqrt(max(1.0, N_d / N_0))
args.width = int(round(args.width * s_scale))
args.height = int(round(args.height * s_scale))
out_root = args.output_root
img_dir = out_root / "images"
img_dir.mkdir(parents=True, exist_ok=True)
rng = random.Random(args.seed)
records = []
sys.path.insert(0, str(Path(__file__).resolve().parents[3]))
from _sample_pool import parallel_sample_records # noqa: E402
def _attempt(rng):
return sample_instance(
rng, args.width, args.height,
min_marks_total=args.min_marks,
max_marks_total=args.max_marks,
min_crossings=args.min_crossings,
max_crossings=args.max_crossings,
min_pixel_between_marks=min_pixel_between_marks,
s_endpoint_excl_px=s_endpoint_excl_px,
num_waypoints=num_waypoints,
num_extra_distractor_curves=num_extra_distractor_curves,
max_attempts=50,
)
records_raw = parallel_sample_records(
_attempt, count=args.count, workers=args.workers,
seed_base=args.seed,
)
rng_render = random.Random(args.seed ^ 0xA5A5)
for idx, record in enumerate(records_raw):
name = f"traverse_ordering_{idx:05d}.png"
ns = rng_render.randint(0, 10**9)
render_instance(img_dir / name, record, noise_seed=ns)
polylines = record.pop("polylines")
record["polylines"] = [_polyline_to_list(p) for p in polylines]
record["image"] = f"images/{name}"
records.append(record)
print(f" {len(records)}/{args.count} valid samples (workers={args.workers})")
with (out_root / "annotations.jsonl").open("w") as fh:
for r in records:
fh.write(json.dumps(r) + "\n")
data_json = {
"task": "traverse_ordering",
"category": "sequential_traversal",
"count": len(records),
"items": [
{"image": r["image"], "question": r["question"], "answer": r["answer"]}
for r in records
],
}
(out_root / "data.json").write_text(json.dumps(data_json, indent=2))
print(f"Saved {len(records)} items to {out_root}")
if __name__ == "__main__":
main()
|