Groups and subclusters of galaxies are formed in a process of merging of individual galaxies. It is hard to define a spatial group member galaxy because of various factors: (1) redshift space distortions; (2) only half of the 6-dimensional spatial-velocity phase space information is known, namely two sky coordinates and redshift of each galaxy, i.e. only partly known kinematics and dynamics; (3) limited observations from only one point in the Universe; etc. Assignment of galaxies to a specific group or subcluster we call a problem of substructure. For solving it there is a need for some non-subjective model independent approach, like the spline method, that includes the construction of a spline histogram and various statistical tests (e.g. Kolmogorov-Smirnov analysis). The spline histogram is a non-parametric representation of a continuous probability density function of a given discreet data set. Generally to solve the problem of substructure, one needs to use various wavelenght information, (e.g. optical, including photometry, x-rays, radio). For more information and overview over different methods of substructure analysis see my Master Thesis "Substructure of Clusters of Galaxies: Methodology" (gziped postscript file, ~1Mb) All these methods may be used in solving various statistical type problems of multiple component data.