Skip to main content
. 2013 Sep 30;3(12):2105–2114. doi: 10.1534/g3.113.007807

Table 5. Accuracy of predictions between different environments.

SR_FI2011 SR_MWS2011 C_SWS2012 SR_FI2012 SR_MWS2012
1 0.292 0.319 0.263 0.294 0.414 0.405
2 0.221 0.234 0.192 0.205 0.626 0.637
3 0.291 0.312 0.251 0.275 0.761 0.760
4 0.569 0.641 0.319 0.310 0.592 0.624
5 0.626 0.681 0.258 0.249 0.622 0.619
6 0.628 0.718 0.403 0.426 0.458 0.453
7 0.560 0.639 0.329 0.326 0.592 0.620
8 0.604 0.662 0.271 0.269 0.610 0.615
9 0.624 0.693 0.430 0.445 0.466 0.465
10 0.088 0.109 0.330 0.358 0.303 0.325

In each case (1−10), two environments were used to train the prediction model. SR_FI2011, Santa Rosa Full irrigated in 2011; SR_MWS2011, Santa Rosa mild water stress in 2011; C_SWS2012, Cauqenes severe water stress in 2012; SR_FI2012, Santa Rosa full irrigated in 2012; SR_MWS2012,Santa Rosa mild water stress 2012.

The training sets were 1: SR_FI2012/SR_MWS2012; 2: C_SWS2012/SR_MWS2012; 3: C_SWS2012/SR_FI2012; 4: SR_MWS2011/ SR_MWS2012; 5: SR_MWS2011/SR_FI2012; 6: SR_MWS2011/C_SWS2012; 7: SR_FI2011/SR_MWS2012, 8: SR_FI2011/SR_FI2012; 9: SR_FI2011/C_SWS2012; 10: SR_FI2011/SR_MWS2011.