Bartley David L, Ogden Trevor, Song Ruiguang
National Institute for Occupational Safety and Health, 4676 Columbia Parkway, Cincinnati, OH 45226, USA.
Biosystems. 2002 Aug-Sep;66(3):179-91. doi: 10.1016/s0303-2647(02)00053-9.
The time-dependent frequency distribution of groups of individuals versus group size was investigated within a continuum approximation, assuming a simplified individual growth, death and creation model. The analogy of the system to a physical fluid exhibiting both convection and diffusion was exploited in obtaining various solutions to the distribution equation. A general solution was approximated through the application of a Green's function. More specific exact solutions were also found to be useful. The solutions were continually checked against the continuum approximation through extensive simulation of the discrete system. Over limited ranges of group size, the frequency distributions were shown to closely exhibit a power-law dependence on group size, as found in many realizations of this type of system, ranging from colonies of mutated bacteria to the distribution of surnames in a given population. As an example, the modeled distributions were successfully fit to the distribution of surnames in several countries by adjusting the parameters specifying growth, death and creation rates.
在连续近似的情况下,假设一个简化的个体生长、死亡和产生模型,研究了个体群体随时间变化的频率分布与群体规模之间的关系。利用该系统与表现出对流和扩散的物理流体的类比,获得了分布方程的各种解。通过应用格林函数对一般解进行了近似。还发现更具体的精确解也很有用。通过对离散系统的广泛模拟,不断地根据连续近似对解进行检验。在有限的群体规模范围内,频率分布被证明与群体规模密切呈现幂律依赖关系,正如在这类系统的许多实例中所发现的那样,从突变细菌菌落到给定人群中的姓氏分布。例如,通过调整指定生长、死亡和产生率的参数,将模拟分布成功地拟合到了几个国家的姓氏分布。