饱和试验设计资料的逐步回归分析

对饱和试验设计资料,一次最小二乘分析是不合理的,采用逐步回归分析法,在选择适当的理论F值前提下,不仅可以检验每个回归系数的显著性,而且能够获得最优回归方程．     

适用于获取最优化配方的一种算法

本文应用印楝种仁提取物（F3）与敌敌畏混配为例,以斜纹夜蛾（Spodopteralitura）为目标害虫,介绍一种适用于获取最优化配方的算法,在二次通用回归旋转组合设计的基础上,经参数辨识,获取二次回归方程,经失拟性、回归显著性检验,本方程基本能够反映杀虫剂用量与斜纹夜蛾幼虫死亡机率值之间的关系．在害虫防治实践中,要求在防治费用最小的基础上,目标害虫有最大的死亡率.因此,以防治目标害虫的费用作为优化算法的目标函数,以害虫死亡机率值最大作为约束条件,有如下的一组优化算式为目标函数约束条件式中a1,a2分别为参试杀虫剂1,2最低用量,b1,b2则为相应的最高用量．C1,C2分别为杀虫剂1,2的单价,N1,N2为杀虫剂l,2的用量．Y为目标害虫死亡机率值回归方程．本文所依据的试验设计中,以对数函数关系变换编码值与使用浓度之间的关系,所以应用拉格朗日求极值原理求取最优化配方．由计算所得的混配比例与其他方法所获结果一致．     

Statistical analysis of algal and water quality data

Data on cyanobacteria (blue-green algae) are generallycollected on a reactive basis, frequently in responseto bloom events. Such data presents a biased andincomplete snapshot of water quality. This paper looksat two typical data sets for UK waters showing thatwhile statistics may be used to describe the data theyare of limited use in forecasting. Suggestions ofappropriate tests for small and sparse data sets aremade.     

昆虫发育进度的非正态描述──分段指数曲线

提出了分段连续指数曲线转化为多元线性回归方程的方法。利用取剩余平方和最小的方法确定未知折点。给出了在描述昆虫发育进度的模拟模型。     

Myocardial uptake kinetics of tocainide enantiomers in the isolated perfused rabbit heart

The extent of myocardial accumulation of tocainide, administered as single enantiomers and as well as racemate, was determined in the isolated, spontaneous beating rabbit heart. The heart was retrogradely perfused at a constant rate and fractions of the perfusate were collected during and after infusion. Kinetic parameters for myocardial accumulation and disposition of tocainide were indirectly determined from drug concentration/time course in the outflow perfusate. No stereoselectivity in myocardial accumulation was observed. A two compartment model with mean half-lives for distribution and elimination of 0.60 and 3.78 min, respectively, was fitted to the accumulation and disposition data. At steady-state, tocainide enantiomers were accumulated about three times in the myocardium relative to the perfusion liquid. © 1995 Wiley-Liss, Inc.