Other Solvers

,

Method	CDF (for Figure 3.2, L)	Noise Histogram	Figure 3.1	Figure 4.1 (L)	Figure 4.1 (R)	Figure 4.1 (R) for Intlab	Method	Figure 4.2	Figure 5.1	Figure 6.1	Method
pcg							pcg				pcg
symmlq							symmlq				symmlq
bicgstab							bicgstab				bicgstab
idr							idr				idr
gmres							gmres				gmres
minres							minres				minres
minresqlprr							minresqlprr				minresqlprr
Method	CDF (for Figure 3.2, L)	Noise Histogram	Figure 3.1	Figure 4.1 (R)	Method	Figure 4.2	Figure 5.1	Figure 6.1	Method

Sec		pcg	symmlq	bicgstab	idr	gmres	minres	minresqlprr
3	Krylov convergence in ECNoise	438	439	435	406	450	446	446
3	ECnoise returns noise estimate	435	439	431	399	450	446	446
3	% probs. with noise estimated	99.3	100.0	99.1	98.3	100.0	100.0	100.0
3	f % with noise > 1e-10	87.7	86.7	89.9	79.8	46.6	87.8	76.8
4	% reAD/noise in [0.01,100]	83.6	76.2	80.8	82.1	88.4	80.1	83.3
4	% dnoise/noise in [0.1,10]	71.3	80.8	80.5	47.2	81.8	81.6	83.3
4	% dnoise >= 0.1*noise	98.1	97.1	100.0	100.0	100.0	98.1	98.8
5	Krylov convergence in f2est	382	382	368	333	385	377	378
5	f2est estimates hopt	377	380	366	330	383	373	374
5	% reFDAD/reFD in [0.1,10]	91.6	92.1	90.2	92.5	94.7	90.3	89.7
6	% of noise <= 1e-10 with reAD <= 0.05	99.4	100.0	100.0	100.0	97.9	100.0	100.0
6	% of noise <= 1e-10 with reFD <= 0.05	96.1	97.1	98.8	98.8	90.1	96.3	96.5
6	% of noise <= 1e-10 and Krylov convergence	96.5	97.4	99.6	99.6	90.1	96.3	96.5
6	% of noise >= 1e-10 and Krylov convergence	64.0	62.6	66.5	57.5	58.5	54.4	55.8
6	% noise >= 1e-10 and Krylov convergence with hopt > 0	98.8	98.8	99.2	98.9	94.7	94.1	94.4
6	% noise >= 1e-10 and Krylov convergence with reAD <= 0.05	98.8	96.3	30.9	65.9	97.4	100.0	98.6
6	% noise >= 1e-10 and Krylov convergence with reFD <= 0.05	90.0	86.6	61.0	79.5	94.7	89.7	90.3