Pt heptamer island

SystemMorse Pt

Atoms343

Runs59

Cap3000 force calls

The main metric is Force calls. Scoring rules.

Charts

Success rate on Pt heptamer island, in rank order. The rate is the share of runs that finished. 10 ranked rows did not record a success rate and are not drawn. PDF.

Dolan-Moré profile for Band / NEB on Pt heptamer island. The curve is the fraction of starts whose cost is at most tau times the best finite cost in this family. Tau is 1, 2, 4, 8, and 16. An unsolved start stays in the denominator and does not cross a finite tau. Dolan, E. D. and Moré, J. J., Math. Program. 91:201-213, 2002. Colored lines are the six largest areas. The other curves are gray. PDF.

Dolan-Moré profile for Minimum mode on Pt heptamer island. The curve is the fraction of starts whose cost is at most tau times the best finite cost in this family. Tau is 1, 2, 4, 8, and 16. An unsolved start stays in the denominator and does not cross a finite tau. Dolan, E. D. and Moré, J. J., Math. Program. 91:201-213, 2002. PDF.

Dolan-Moré profile for Connect / double-ended on Pt heptamer island. The curve is the fraction of starts whose cost is at most tau times the best finite cost in this family. Tau is 1, 2, 4, 8, and 16. An unsolved start stays in the denominator and does not cross a finite tau. Dolan, E. D. and Moré, J. J., Math. Program. 91:201-213, 2002. PDF.

Force calls on each recorded start for Pt heptamer island. The box is the quartiles of those counts. A blue point is a success and a vermillion cross is a recorded failure (233). The horizontal position is the recorded count. The vertical offset only keeps marks from covering one another. PDF.

Performance profile. Higher means more starts solved within that cost ratio. Colored lines are the six largest areas. Thin lines are the other 17.

• Eon svn (r2549) · Lanczos improved L-BFGS
• Eon svn (r2549) · CI-NEB
• Eon svn (r2549) · LBFGS
• Eon svn (r2549) · Dimer L-BFGS
• Eon svn (r2549) · Lanczos L-BFGS
• Eon svn (r2549) · Doubly nudged CI-NEB

Cost distribution of 33 ranked rows. Logarithmic bins, from 103 to 6000.

Performance profile SVG Cost and success SVG

Performance profile. Select the plot for the full SVG.

Cost and success. Select the plot for the full SVG.

A shorter bar is a lower cost on a log scale shared by the codes of this problem.

Leaderboard

A shorter bar is a lower cost on the log scale for that problem. Select a column heading to sort.

Rank	Code	Best	Median	Success	PAR-2
1	OPTIM	103	211	100.0%	103
2	Eon svn	110	225.881	100.0%	110
3	pele	449.746	449.746	96.6%	449.746
4	geodesic	824.119	1115.25	98.3%	824.119
5	ASE	926.475	926.475	98.3%	926.475
6	rgsaddle	938.102	1070.44	93.2%	938.102
7	Sella	968.525	968.525	98.3%	968.525
8	COMPASS	1078.12	1078.12	98.3%	1078.12
9	TSASE	2524.53	4262.26	64.4%	2524.53
10	LAMMPS	2668.64	3617.71	74.6%	2668.64
11	eOn	3080.19	3080.19	64.4%	3080.19

The tables below list every ranked row, by method family and overall.

eOn band rows are not yet on the harness counter.

A run short of the full set, or a run that is not the stock method, stays off the ranked table.

These archived rows differ from the table of J. Chem. Theory Comput. 10, 5476 (2014) that prints them, and each ranks on its archived values: eOn (r2156) · eon-lanczos (Table 10 prints 165, 72 and 308 calls with 0 failed, this file has 325, 95 and 2660 with 5 failed, and no archived file has the printed values. It was written on 29 Aug 2013, six days after the files that match the other eOn rows of Table 10).

A failure class marked inferred comes from a start file that records no class. A start with an error is crash, a start at or above the cap is cap, and any other failure is divergence unless its force norm is at or below the stop, which makes it wrong_stationary. Rows with an inferred class: Sella (2.6.0) · Sella+NEB.

Band / NEB

Band / NEB. One row is one code, one version, and one method. A lower cost is better. PAR-2 counts a finished run as the calls it used and a failed run as twice the cap. An archived row shows the average it published, and its PAR-2 applies that rule to the failures the entry records. Hessians is the mean analytic Hessian count. FD calls is that count times one force call per coordinate.

Rank	Row	Cost	Unit	Hessians	FD calls	Runs
1	eOn (r2156) · eon-ci-neb-single	110	force calls	0	0	59
2	Eon svn (r2549) · CI-NEB	225.881	force calls	0	0	59
3	Eon svn (r2549) · Doubly nudged CI-NEB	246.271	force calls	0	0	59
4	eOn (r2156) · eon-ci-neb	248	force calls	0	0	59
5	eOn (r2156) · eon-ci-swdneb	256	force calls	0	0	59
6	eOn (r2156) · eon-ci-neb-5	391	force calls	0	0	59
7	geodesic (1.0.0) · NEB-FIRE geodesic	824.119	force calls	0	0	59
8	ASE (3.29.0) · NEB-FIRE	926.475	force calls	0	0	59
9	rgsaddle (0.2.2) · CI-NEB energy-weighted	938.102	force calls	0	0	59
10	Sella (2.6.0) · Sella+NEB	968.525	force calls	0	0	59
11	rgsaddle (0.2.2) · OCI-NEB	975.983	force calls	0	0	59
12	rgsaddle (0.2.2) · IDPP CI-NEB	1062.46	force calls	0	0	59
13	rgsaddle (0.2.2) · CI-NEB	1078.42	force calls	0	0	59
14	geodesic (1.0.0) · CI-NEB-LBFGS geodesic	1115.25	force calls	0	0	59
15	rgsaddle (0.2.2) · S-IDPP CI-NEB	1134.93	force calls	0	0	59
16	geodesic (1.0.0) · CI-NEB geodesic	1462.14	force calls	0	0	59
17	TSASE (1.0) · CI-NEB	2524.53	force calls	0	0	59
18	LAMMPS (30 Sep 2026) · CI-NEB-fire	2668.64	force calls	0	0	59
19	LAMMPS (30 Sep 2026) · CI-NEB-quickmin	4566.78	force calls	0	0	59
20	TSASE (1.0) · NEB	6000	force calls	0	0	59
21	rgsaddle (0.2.2) · NEB	6000	force calls	0	0	59

Minimum mode

Minimum mode. One row is one code, one version, and one method. A lower cost is better. PAR-2 counts a finished run as the calls it used and a failed run as twice the cap. An archived row shows the average it published, and its PAR-2 applies that rule to the failures the entry records. Hessians is the mean analytic Hessian count. FD calls is that count times one force call per coordinate.

Rank	Row	Cost	Unit	Hessians	FD calls	Runs
1	Eon svn (r2549) · Lanczos improved L-BFGS	180.644	force calls	0	0	59
2	eOn (r2156) · eon-dimer	182	force calls	0	0	59
3	Eon svn (r2549) · Dimer L-BFGS	192.051	force calls	0	0	59
4	Eon svn (r2549) · Lanczos L-BFGS	221.525	force calls	0	0	59
5	eOn (r2156) · eon-lanczos	805.932	force calls	0	0	59
6	eOn (r2156) · eon-lanczos-improved	1005.46	force calls	0	0	59

Connect / double-ended

Connect / double-ended. One row is one code, one version, and one method. A lower cost is better. PAR-2 counts a finished run as the calls it used and a failed run as twice the cap. An archived row shows the average it published, and its PAR-2 applies that rule to the failures the entry records. Hessians is the mean analytic Hessian count. FD calls is that count times one force call per coordinate.

Rank	Row	Cost	Unit	Hessians	FD calls	Runs
1	OPTIM · DNEB + hybrid eigenvector following	103	force calls	0	0	59
2	Eon svn (r2549) · LBFGS	225.881	force calls	0	0	59
3	OPTIM · climbing string	319	force calls	0	0	59
4	pele · hybrid eigenvector following	449.746	force calls	0	0	59
5	COMPASS (1.0.0) · COMPASS	1078.12	force calls	0	0	59
6	eOn (3.6.0) · LBFGS	3080.19	force calls	0	0	59

Overall

One row is one code, one version, and one method. A lower cost is better. PAR-2 counts a finished run as the calls it used and a failed run as twice the cap. An archived row shows the average it published, and its PAR-2 applies that rule to the failures the entry records. Hessians is the mean analytic Hessian count. FD calls is that count times one force call per coordinate. The per-image mean divides the total by the number of moving images.

Row	Family	Family rank	Unit	PAR-2	Published	Median calls	Mean calls	Per-image mean	Hessians	FD calls	Failures	Runs
OPTIM · DNEB + hybrid eigenvector following	Connect / double-ended	1	force calls	103	103	98	103		0	0	0 failed	59
eOn (r2156) · eon-ci-neb-single	Band / NEB	1	force calls	110	110	62	110		0	0	0 failed	59
Eon svn (r2549) · Lanczos improved L-BFGS	Minimum mode	1	force calls	180.644		161	180.644	60.2147	0	0	crash 0, divergence 0, cap 0, wrong_stationary 0	59
eOn (r2156) · eon-dimer	Minimum mode	2	force calls	182	182	182	182		0	0	0 failed	59
Eon svn (r2549) · Dimer L-BFGS	Minimum mode	3	force calls	192.051		188	192.051	64.0169	0	0	crash 0, divergence 0, cap 0, wrong_stationary 0	59
Eon svn (r2549) · Lanczos L-BFGS	Minimum mode	4	force calls	221.525		208	221.525	73.8418	0	0	crash 0, divergence 0, cap 0, wrong_stationary 0	59
Eon svn (r2549) · CI-NEB	Band / NEB	2	force calls	225.881		188	225.881	75.2938	0	0	crash 0, divergence 0, cap 0, wrong_stationary 0	59
Eon svn (r2549) · LBFGS	Connect / double-ended	2	force calls	225.881		188	225.881	75.2938	0	0	crash 0, divergence 0, cap 0, wrong_stationary 0	59
Eon svn (r2549) · Doubly nudged CI-NEB	Band / NEB	3	force calls	246.271		218	246.271	82.0904	0	0	crash 0, divergence 0, cap 0, wrong_stationary 0	59
eOn (r2156) · eon-ci-neb	Band / NEB	4	force calls	248	248	211	248		0	0	0 failed	59
eOn (r2156) · eon-ci-swdneb	Band / NEB	5	force calls	256	256	219	256		0	0	0 failed	59
OPTIM · climbing string	Connect / double-ended	3	force calls	319	319	286	319		0	0	0 failed	59
eOn (r2156) · eon-ci-neb-5	Band / NEB	6	force calls	391	391	395	391		0	0	0 failed	59
pele · hybrid eigenvector following	Connect / double-ended	4	force calls	449.746	255	244	255		0	0	2 failed, class not recorded	59
eOn (r2156) · eon-lanczos	Minimum mode	5	force calls	805.932	325	207.5	325		0	0	5 failed, class not recorded	59
geodesic (1.0.0) · NEB-FIRE geodesic	Band / NEB	7	force calls	824.119		716	773.271	257.757	0	0	crash 0, divergence 0, cap 1, wrong_stationary 0	59
ASE (3.29.0) · NEB-FIRE	Band / NEB	8	force calls	926.475		707	875.712	291.904	0	0	crash 0, divergence 1, cap 0, wrong_stationary 0	59
rgsaddle (0.2.2) · CI-NEB energy-weighted	Band / NEB	9	force calls	938.102		478	561.644	187.215	0	0	crash 0, divergence 2, cap 0, wrong_stationary 2	59
Sella (2.6.0) · Sella+NEB	Band / NEB	10	force calls	968.525		756	921.847	307.282	0	0	crash 0, divergence 0, cap 1 (1 inferred), wrong_stationary 0	59
rgsaddle (0.2.2) · OCI-NEB	Band / NEB	11	force calls	975.983		602	823.441	274.48	0	0	crash 0, divergence 0, cap 3, wrong_stationary 0	59
eOn (r2156) · eon-lanczos-improved	Minimum mode	6	force calls	1005.46	222	157	222		0	0	8 failed, class not recorded	59
rgsaddle (0.2.2) · IDPP CI-NEB	Band / NEB	12	force calls	1062.46		491	585.729	195.243	0	0	crash 0, divergence 5, cap 0, wrong_stationary 0	59
COMPASS (1.0.0) · COMPASS	Connect / double-ended	5	force calls	1078.12		1002	1027.27	342.424	0	0	crash 0, divergence 0, cap 1, wrong_stationary 0	59
rgsaddle (0.2.2) · CI-NEB	Band / NEB	13	force calls	1078.42		479	601.542	200.514	0	0	crash 0, divergence 5, cap 0, wrong_stationary 0	59
geodesic (1.0.0) · CI-NEB-LBFGS geodesic	Band / NEB	14	force calls	1115.25		209	223.847	74.6158	0	0	crash 0, divergence 9, cap 0, wrong_stationary 0	59
rgsaddle (0.2.2) · S-IDPP CI-NEB	Band / NEB	15	force calls	1134.93		461	607.153	202.384	0	0	crash 0, divergence 5, cap 1, wrong_stationary 0	59
geodesic (1.0.0) · CI-NEB geodesic	Band / NEB	16	force calls	1462.14		491	604.746	201.582	0	0	crash 0, divergence 9, cap 0, wrong_stationary 0	59
TSASE (1.0) · CI-NEB	Band / NEB	17	force calls	2524.53		564	639.61	213.203	0	0	crash 0, divergence 20, cap 1, wrong_stationary 0	59
LAMMPS (30 Sep 2026) · CI-NEB-fire	Band / NEB	18	force calls	2668.64		1405	1471.69	490.565	0	0	crash 0, divergence 6, cap 0, wrong_stationary 9	59
eOn (3.6.0) · LBFGS	Connect / double-ended	6	force calls	3080.19		2260	3220.9	1073.63	0	0	crash 0, divergence 0, cap 21, wrong_stationary 0	59
LAMMPS (30 Sep 2026) · CI-NEB-quickmin	Band / NEB	19	force calls	4566.78		2760	2508.98	836.328	0	0	crash 0, divergence 26, cap 0, wrong_stationary 9	59
TSASE (1.0) · NEB	Band / NEB	20	force calls	6000		543	788.746	262.915	0	0	crash 0, divergence 56, cap 3, wrong_stationary 0	59
rgsaddle (0.2.2) · NEB	Band / NEB	21	force calls	6000		533	744.475	248.158	0	0	crash 0, divergence 56, cap 3, wrong_stationary 0	59

Off the ranked table

• eOn (3.6.0) · LBFGS n=59, not yet on the harness counter.
• eOn (3.6.0) · LBFGS n=59, not yet on the harness counter.