The least-squares characteristics related to the regression of sorghum production onto population, precipitation, and population and precipitation are presented in Table 5.
Table 5. Least squares statistics associated with of the regressions of Maize
3.3.4. Rice
The analysis of rice yield in MATLAB revealed again that the sum of sine model of three terms was superior to other statistical models such as regression polynomials, Gaussian and rational models (Figure 7.a). Its adjusted R2 0.8025 with an RMSE of 2.33x104. We suggest that the rice production is more precipitation- dependent than population-dependent. This is evident from the following two model fits of rice and precipitation (See Figure 7). As shown in Figure 7a and 7b, there is clear synchrony between the two model fits. In particular, both models indicate a dip at year 5 and a spike around year 32. The variations in between are also in good agreement (see supplemental V for the details of the model fit).