Table 2 Regression models to obtain the mean diameter, \(\overline{D}\) (cm), and the dominant diameter \({D}_{0}\) (cm). The term \(\varepsilon\) represents a generic error term in each regression model. Coef std err: coefficient standard error. Res std err: residual standard error. Res std err CV and \({R}^{2}\) CV are the residual standard error and coefficient of determination obtained using leave-one-out cross-validation, respectively. \({D}_{\mathrm{g}}\) (cm) is the quadratic mean diameter and \(N\) the stand density (trees ha−1)
Model | \(\overline{D}=a+b{D}_{\mathrm{g}}+\varepsilon\) | ||||||||
|---|---|---|---|---|---|---|---|---|---|
Coefficient | Estimate | Coef std err | t value | p value | \({R}^{2}\) | \({R}^{2}\) CV | Res std err | Res std err CV | |
a | − 6.23E-1 | 6.35E-2 | − 9.95 | < 1E-5 | 99.95% | 99.95% | 0.23 cm | 0.24 cm | |
b | 9.96E-1 | 2.10E-3 | 474.8 | < 1E-5 | |||||
Model | \({D}_{\mathrm{o}}=a+b{D}_{\mathrm{g}}+c\frac{1}{N}+\varepsilon\) | ||||||||
Coefficient | Estimate | Coef std err | t value | p value | \({R}^{2}\) | \({R}^{2}\) CV | Res std err | Res std err CV | |
a | 3.18 | 6.21E-1 | 5.12 | < 1.E-5 | 97.99% | 97.87% | 1.75 cm | 1.80 cm | |
b | 1.30 | 3.81E-2 | 34.11 | < 1E-5 | |||||
c | − 3017 | 544 | 5.54 | < 1E-5 | |||||