Table 2.
Full models precision and accuracy metrics for the RAU, yield, and protein concentration.
| a) Model metrics | |||||
| Model | MAE | RMSE | RRMSE | R2 | |
| RAU (%) | Null | 6.1 (±2.8)1 | 7.0 (±3.3) | 8.1 (±4.6) | |
| Full | 3.4 (±1.5) | 3.9 (±1.6) | 4.5 (±2.2) | 0.66 (±0.3) | |
| Yield (Mg ha−1) | Null | 0.4 (±0.1) | 0.5 (±0.2) | 11.9 (±4.2) | |
| Full | 0.2 (±0.1) | 0.3 (±0.1) | 6.9 (±3.2) | 0.68 (±0.3) | |
| Protein (g kg−1) | Null | 10 (±3.3) | 12 (±3.8) | 2.9 (±0.9) | |
| Full | 4.8 (±1.6) | 5.45 (±1.8) | 1.4 (±0.4) | 0.70 (±0.3) | |
| b) Random effects variance | |||||
| Environment | Residual | ||||
| RAU | Null | 31.5 (±3.1) | 26.6 (±1.4) | ||
| Full | 0 (±0) | 9.2 (±0.5) | |||
| Yield | Null | 0.10 (±0.01) | 0.13 (±0.01) | ||
| Full | 0.001 (±0) | 0.01 (±0.01) | |||
| Protein | Null | 118.5 (±4.6) | 56.5 (±1.7) | ||
| Full | 0 (−±0) | 16.1 (±0.7) | |||
Mean absolute error (MAE), root mean squared error (RMSE), relative root mean squared error (RRMSE), and coefficient of determination (R2). (a), Full model variance partition; (b) Random effects variance.
1
Median value followed by the standard deviation from the cross-validation procedure.