为了用线性或非线性模型拟合纵向数据，需要确定一个合理的协方差结构。
It is necessary to make a reasonable covariance structure for longitudinal data so as to fit these data with linear or nonlinear models.
本文根据以前传统或复合基团相互作用法的模型并且考虑到本方法模指数的概念，提出了一般物性估算模型方程：，根据此模型对1228种物质进行拟合得出基团相互作用键的贡献值。
In view of simple group contribution and bond module index, a universal model was set up as the following:, and bond contribution data from 1228 substances were calculated by this equation.
传统逆向工程要经过曲面拟合、等距、裁剪或三角网格剖分建立其曲面模型，进而才能在曲面模型上生成数控加工刀位。
For the traditional reverse engineering, the tool-path is generated based on the surface fitting, surface offset and triangulation to construct the surface model.
然而，由于此类模型的特殊性，我们不能利用一些通常的统计软件对它进行模型拟合或参数估计。
But it can not be fitted directly by using some general statistical software, due to its speciality. For this problem, this paper propose two methods which are both feasible and convenient.
结果表明：12个农艺性状的遗传基本符合双基因工作六参数模型或部分拟合模型。
The results showed the genetic model of 12 agronomic traits were fundamentally fitted for 6-parameter model.
结果表明：12个农艺性状的遗传基本符合双基因工作六参数模型或部分拟合模型。
The results showed the genetic model of 12 agronomic traits were fundamentally fitted for 6-parameter model.
应用推荐