Table 3. Accuracy of predictions for each trial in 2011 using random training sets with 100 independent randomizations.

			IB	RC	RCB_MVNG	MVNG
SR_FI	GY	RR	0.298 ± 0.117	0.296 ± 0.119	0.319 ± 0.114	0.319 ± 0.113
		GAUSS	0.312 ± 0.117	0.310 ± 0.120	0.325 ± 0.117	0.326 ± 0.116
	TKW	RR	0.780 ± 0.056	0.780 ± 0.056	0.777 ± 0.057	0.843 ± 0.040
		GAUSS	0.786 ± 0.055	0.786 ± 0.055	0.782 ± 0.056	0.847 ± 0.039
	DH	RR	0.409 ± 0.109	0.409 ± 0.109	0.405 ± 0.109	0.579 ± 0.123
		GAUSS	0.436 ± 0.111	0.436 ± 0.111	0.433 ± 0.111	0.614 ± 0.121
	NKS	RR	0.479 ± 0.114	0.479 ± 0.114	0.484 ± 0.115	0.665 ± 0.077
		GAUSS	0.487 ± 0.119	0.487 ± 0.119	0.492 ± 0.120	0.669 ± 0.075
SR_MWS	GY	RR	0.236 ± 0.141	0.275 ± 0.147	0.231 ± 0.127	0.347 ± 0.134
		GAUSS	0.231 ± 0.144	0.273 ± 0.150	0.260 ± 0.128	0.370 ± 0.132
	TKW	RR	0.759 ± 0.061	0.762 ± 0.061	0.757 ± 0.058	0.841 ± 0.034
		GAUSS	0.764 ± 0.059	0.767 ± 0.059	0.761 ± 0.057	0.845 ± 0.034
	DH	RR	0.398 ± 0.110	0.399 ± 0.110	0.396 ± 0.110	0.563 ± 0.134
		GAUSS	0.423 ± 0.108	0.423 ± 0.108	0.423 ± 0.108	0.604 ± 0.134
	NKS	RR	0.464 ± 0.115	0.466 ± 0.114	0.458 ± 0.114	0.608 ± 0.088
		GAUSS	0.483 ± 0.111	0.485 ± 0.111	0.478 ± 0.111	0.608 ± 0.086

IB, incomplete blocks, field design; RC, row by column model; RCB_MVNG, random complete block model with moving means as covariable; MVNG, linear regression model with moving means as covariable; SR_FI, Santa Rosa under full irrigation; GY, grain yield; RR, Ridge regression kernel; GAUSS, Gaussian kernel; TKW, thousand kernel weight; DH, days to heading; NKS, number of kernels per spike; SR_MWS, Santa Rosa under mild water stress.