vvs/sim-numerics
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
Files
1842eb07659c1dbdf30dfa39b10b0e63bbe952cd
sim-numerics/results_diff_signfix.txt
T
vvsandClaude Opus 5 68df1d25db Sign tests by product underflowed: FIXED everywhere (branch, not master - the paper is not yet updated). FA * fnew < 0.0 with FA ~ -1e-200, fnew ~ 1e-150 is -0.0 and '-0.0 < 0.0' is False, so on flat functions (x^19) the bracket update moved the wrong end, the root left the bracket and the run still reported convergence - the proposed rule returned 4.9e-9, bisection 3.5e-9, ITP 1.6e-8 with a claimed 1e-12 bracket; the 1e-6 harness scored them solved. Same defect in our bisection, false-position family, Ridders, Brent, ITP and twelve copies in stage scripts, simroot/sim.py and one test; all replaced by _opposite / _same_sign, pinned by test_sign_underflow. 
Impact, every sweep regenerated: the paper's key configurations are unchanged to the evaluation (proposed 2806/3220, credit 3568/3020, Chandrupatla, TOMS 748, Brent, bisection); 46 stated numbers move (ITP CHA 2857 -> 2896, ITP oracle 2611 -> 2647, record CHA figures by 14-18, the grid means by ~1%) and one finding weakens: the measured-order exponent of section 7.1 was cheaper in 29 of 60 far cells and 7 of 7 safeguards; with the fix 12 of 61 and 6 of 7, and on the wide-bracket set it now loses in every safeguard and drops 5 problems.

PREREG G1-G3 outcome on the fixed code: G2 (repeating step pattern, period from the data) PASSES Y1-Y4 and is ACCEPTED by the stop rule - proposed rule APS 2806 -> 2782 (below Chandrupatla's 2787), CHA 3220 -> 3169, both 6026 -> 5951, per problem vs Chandrupatla on APS 95/20/38 p = 8e-7. G1 fails Y1/Y2 (its Y3 miss was the sign bug), G3 fails Y2, A fails Y2. Regression 887 pass; the two red tests are the paper-number check (awaiting the paper update) and test_blend (fixed since, copy in stage16).

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-09-21 12:49:01 +01:00

79 lines
6.6 KiB
Plaintext
Raw Blame History

CHANGED pure.ratio.derived3.step CHA: total 3460 -> 3524, solved 40 -> 40, over2x 10 -> 15
CHANGED pure.ratio.start.both CHA: total 3631 -> 3626, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio.start.project CHA: total 5029 -> 5024, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.bracket.both CHA: total 3315 -> 3335, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.bracket.bracket CHA: total 3321 -> 3341, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.bracket.credit CHA: total 3130 -> 3129, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.bracket.project CHA: total 4763 -> 4763, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.bracket.safe CHA: total 3998 -> 4059, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.bracket.step CHA: total 4004 -> 4065, solved 40 -> 40, over2x 2 -> 2
CHANGED pure.ratio_soft.derived.both CHA: total 3308 -> 3358, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.derived.bracket CHA: total 3167 -> 3235, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.derived.credit CHA: total 3094 -> 3094, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.derived.safe CHA: total 3643 -> 3695, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.derived.step CHA: total 3704 -> 3756, solved 40 -> 40, over2x 8 -> 8
CHANGED pure.ratio_soft.derived3.both CHA: total 3370 -> 3435, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.derived3.bracket CHA: total 3513 -> 3581, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.derived3.none CHA: total 4285 -> 4248, solved 35 -> 35, over2x 5 -> 5
CHANGED pure.ratio_soft.derived3.safe CHA: total 3681 -> 3741, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.derived3.step CHA: total 3724 -> 3819, solved 40 -> 40, over2x 5 -> 10
CHANGED pure.ratio_soft.start.both CHA: total 3508 -> 3508, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.ratio_soft.start.project CHA: total 4954 -> 4949, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.rational.bracket.both CHA: total 3314 -> 3323, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.rational.bracket.bracket CHA: total 3433 -> 3440, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.rational.bracket.credit CHA: total 3040 -> 3039, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.rational.bracket.project CHA: total 4676 -> 4671, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.rational.bracket.safe CHA: total 3835 -> 3829, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.rational.bracket.step CHA: total 3841 -> 3835, solved 40 -> 40, over2x 3 -> 3
CHANGED pure.rational.derived.both CHA: total 3674 -> 3715, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.rational.derived.bracket CHA: total 3766 -> 3821, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.rational.derived.safe CHA: total 3557 -> 3637, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.rational.derived.step CHA: total 3618 -> 3698, solved 40 -> 40, over2x 10 -> 10
CHANGED pure.rational.derived3.both CHA: total 3815 -> 3873, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.rational.derived3.bracket CHA: total 3994 -> 4052, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.rational.derived3.safe CHA: total 3505 -> 3538, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.rational.derived3.step CHA: total 3618 -> 3698, solved 40 -> 40, over2x 9 -> 13
CHANGED pure.rational.start.project CHA: total 4762 -> 4758, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.bracket.both CHA: total 3391 -> 3431, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.bracket.bracket CHA: total 3387 -> 3443, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.bracket.safe CHA: total 3989 -> 4038, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.bracket.step CHA: total 4017 -> 4066, solved 40 -> 40, over2x 1 -> 1
CHANGED pure.sqrt.derived.both CHA: total 3870 -> 3908, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.derived.bracket CHA: total 4102 -> 4162, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.derived.none CHA: total 3511 -> 3538, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.derived.project CHA: total 3511 -> 3538, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.derived.safe CHA: total 3493 -> 3520, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.derived.step CHA: total 3493 -> 3525, solved 40 -> 40, over2x 0 -> 1
CHANGED pure.sqrt.derived3.both CHA: total 3737 -> 3781, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.derived3.bracket CHA: total 3859 -> 3904, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.derived3.credit CHA: total 3130 -> 3130, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.derived3.none CHA: total 3852 -> 3883, solved 40 -> 40, over2x 4 -> 4
CHANGED pure.sqrt.derived3.project CHA: total 3642 -> 3649, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.derived3.safe CHA: total 3891 -> 3928, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.derived3.step CHA: total 3924 -> 4039, solved 40 -> 40, over2x 5 -> 10
CHANGED pure.sqrt.start.both CHA: total 3829 -> 3829, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.start.credit CHA: total 3130 -> 3130, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.start.project CHA: total 5120 -> 5115, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.start.safe CHA: total 4395 -> 4395, solved 40 -> 40, over2x 0 -> 0
CHANGED pure.sqrt.start.step CHA: total 4396 -> 4396, solved 40 -> 40, over2x 0 -> 0
CHANGED ridders CHA: total 4216 -> 3768, solved 36 -> 40, over2x 0 -> 0
unexpected changes: 168
rows changed: APS 5 of 264, CHA 163 of 264
label APS old->new CHA old->new solved
pure.ratio.derived3.project 2806->2806 3220->3220 193->193
pure.ratio.derived3.credit 3568->3568 3020->3020 193->193
pure.ratio.derived3.none 2806->2806 4104->4104 188->188
pure.ratio.derived3.safe 3036->3036 3363->3390 193->193
chandrupatla 2787->2787 2441->2441 193->193
toms748 2820->2820 4000->4000 193->193
brent 2889->2889 4115->4115 191->191
bisection 7530->7530 3090->3090 193->193
itp 3957->3957 2857->2896 193->193
itp_oracle 3260->3260 2611->2647 193->193
ridders 3313->3313 4216->3768 189->193
illinois 7045->7045 6234->6234 176->176
regula_falsi 10560->10560 8239->8239 133->133
pure.sqrt.bracket.project 7796->7796 4516->4516 194->194
pure.exp_abs.bracket.both 3461->3461 2733->2736 193->193
proposed rule, CHA problems changed: {}
Reference in New Issue View Git Blame Copy Permalink