Institute for Energy, Environment and Sustainable Communities, University of Regina, Regina, SK, S4S 0A2, Canada.
Environ Sci Pollut Res Int. 2015 Sep;22(18):14220-33. doi: 10.1007/s11356-015-4664-7. Epub 2015 May 14.
This paper presents an open-source software package, rSCA, which is developed based upon a stepwise cluster analysis method and serves as a statistical tool for modeling the relationships between multiple dependent and independent variables. The rSCA package is efficient in dealing with both continuous and discrete variables, as well as nonlinear relationships between the variables. It divides the sample sets of dependent variables into different subsets (or subclusters) through a series of cutting and merging operations based upon the theory of multivariate analysis of variance (MANOVA). The modeling results are given by a cluster tree, which includes both intermediate and leaf subclusters as well as the flow paths from the root of the tree to each leaf subcluster specified by a series of cutting and merging actions. The rSCA package is a handy and easy-to-use tool and is freely available at http://cran.r-project.org/package=rSCA . By applying the developed package to air quality management in an urban environment, we demonstrate its effectiveness in dealing with the complicated relationships among multiple variables in real-world problems.
本文提出了一个开源软件包 rSCA，它是基于逐步聚类分析方法开发的，是用于建立多个因变量和自变量之间关系模型的统计工具。rSCA 包在处理连续和离散变量以及变量之间的非线性关系方面非常有效。它通过一系列的切割和合并操作，根据多元方差分析（MANOVA）的理论，将因变量的样本集分为不同的子集（或子群）。模型结果由一个聚类树给出，该树包括中间和叶子集，以及从树的根到每个叶子集的流路径，这些路径由一系列切割和合并操作指定。rSCA 包是一个方便易用的工具，可在 http://cran.r-project.org/package=rSCA 上免费获取。通过将开发的软件包应用于城市环境中的空气质量管理，我们展示了它在处理实际问题中多个变量之间复杂关系方面的有效性。
