Re-run of the package of the first version by the independent run of the second round (2026-10-14).
Archive: release/source.zip of the first version (sha256 679991d3d9f13776aa61f274271b4fd8a23e164cb70ff5654bc16cb7c6aa1933), extracted
into an empty folder; every program run from its own folder of the copy as 'python3 -I ./PROGRAM ARGS' (Python 3.13.5,
numpy 2.5.2, sympy 1.14.0; nice -n 10; at most three programs at a time; single-threaded numerical libraries).
Comparison: standard output against the recorded output, after removing the line written by the shell's time command
in the recorded outputs and the running-time fields printed by the programs.

automaton_enum_D30: IDENTICAL up to running times; command: python3 -I ./automaton_enum.py 30; exit 0; time 39.2 s; lines 39/39
check_note: IDENTICAL up to running times; command: python3 -I ./check_note.py; exit 0; time 64.41 s; lines 51/51
lamp_fourier_K16: IDENTICAL up to running times; command: python3 -I ./lamp_fourier.py 16; exit 0; time 24.03 s; lines 16/16
lamp_fourier_K22: IDENTICAL up to running times; command: python3 -I ./lamp_fourier.py 22; exit 0; time 520.89 s; lines 22/22
lamp_regular_K8: IDENTICAL up to running times; command: python3 -I ./lamp_regular.py 8; exit 0; time 47.91 s; lines 9/9
lamp_regular_K9: DIFFERENT; command: python3 -I ./lamp_regular.py 9; exit 0; time 752.56 s; lines 10/10
   first differences: [(2, 'K= 2 |G|=      8 dim=     16  dimker mod p: [10]  v_K = 5/4 = 1.250000  predicted 5/4 OK  (T s)', 'K= 2 |G|=      8 dim=     16  dimker mod p: [10]  exact over Q: 10  v_K = 5/4 = 1.250000  predicted 5/4 OK  (T s)'), (3, 'K= 3 |G|=     24 dim=     48  dimker mod p: [8]  v_K = 1/3 = 0.333333  (T s)', 'K= 3 |G|=     24 dim=     48  dimker mod p: [8]  exact over Q: 8  v_K = 1/3 = 0.333333  (T s)'), (4, 'K= 4 |G|=     64 dim=    128  dimker mod p: [24]  v_K = 3/8 = 0.375000  predicted 3/8 OK  (T s)', 'K= 4 |G|=     64 dim=    128  dimker mod p: [24]  exact over Q: 24  v_K = 3/8 = 0.375000  predicted 3/8 OK  (T s)')] normalized lengths 10 10
lamplighter_check: IDENTICAL; command: python3 -I ./lamplighter_check.py 7; exit 0; time 27.4 s; lines 19/18
possible_values_check: IDENTICAL; command: python3 -I ./possible_values_check.py 9; exit 0; time 506.44 s; lines 9/8
selftest_automaton_wrong1: IDENTICAL up to running times; command: python3 -I ./automaton_enum.py 14 --wrong1; exit 0; time 2.22 s; lines 24/24
selftest_automaton_wrong2: IDENTICAL up to running times; command: python3 -I ./automaton_enum.py 14 --wrong2; exit 0; time 0.92 s; lines 23/23
selftest_lamp_regular_wrong: IDENTICAL up to running times; command: python3 -I ./lamp_regular.py 6 --wrong --exact-upto 4; exit 0; time 0.7 s; lines 7/7
vA_automaton: IDENTICAL; command: python3 -I ./vA_automaton.py 26; exit 0; time 6.81 s; lines 18/17
vA_blocks: IDENTICAL; command: python3 -I ./vA_blocks.py 8; exit 0; time 69.11 s; lines 11/10
vA_direct_groupring: IDENTICAL; command: python3 -I ./vA_direct_groupring.py 8 6; exit 0; time 41.69 s; lines 23/22
vA_fourier_model: IDENTICAL; command: python3 -I ./vA_fourier_model.py 17 10; exit 0; time 83.03 s; lines 24/23
vA_fourier_model (command line of the README, without ./): IDENTICAL; command: python3 -I vA_fourier_model.py 17 10; exit 0; time 84.6 s; lines 24/23
verify_construction: IDENTICAL; command: python3 -I ./verify_construction.py 11 4 13; exit 0; time 43.07 s; lines 12/11

Note on lamp_regular_K9.out: the recorded output contains no 'exact over Q' fields; it was produced with the option
--exact-upto 0 (not with the default, which adds exact ranks over Q for K <= 5). With the default command every number
agrees (dimker mod p, v_K, predictions for K = 2..9); with 'lamp_regular.py 8 --exact-upto 0' the lines for K <= 8 are
identical to the recorded ones up to running times (52 s); the K = 9 line agrees with the default re-run (705 s).
Both long runs that had not been repeated at the writing (lamp_regular.py 9, 753 s; lamp_fourier.py 22, 521 s) were
repeated here.
