Broadening classic methods of approximation by implicit functions enables approximation of a set by disconnected figures, e.g. approximation of two or more points simultaneously. As a result it's possible to determine in exact way the 'centers of mass concentration' of an analyzed set - that is to indicate the areas of the set of big congestion of elements. What might be problematic is to determine the proper amount of points to approximate a certain set. How to verify the thesis that two point approximation is more 'proper' than one point approximation? Or to put it in a more general way: how to verify the thesis that it's better to approximate a set by the k-points? In the article the concept of a partial variance is introduced, as a kind of generalization of set variance. The partial variance proves to be a useful tool which enables among others, answering the above questions. It's bound to the method of approximation by disconnected figures, being its valuable completion.
Financed by the National Centre for Research and Development under grant No. SP/I/1/77065/10 by the strategic scientific research and experimental development program:
SYNAT - “Interdisciplinary System for Interactive Scientific and Scientific-Technical Information”.