Talk:Demographic center of Sweden

Conflicting definitions

The source says (on p. 665) that the point was calculated as the intersection of the median latitude and median longitude of the population. This is not the same as the geometric median of the population (which would minimize the cumulative distance, "were they all to meet at a single location"). The geometric median is significantly more difficult to compute; it cannot be found by such a simple trick as taking the coordinatewise medians. — Furthermore, this article claims that the point was computed using a weighted mean, which is yet another point. (Mean is generally different from median, even in one dimension; similarly in two dimensions.) --Jmk (talk) 10:05, 29 September 2009 (UTC)

A more specific source document from SCB echoes this ambiguity. The first description of the center (balance a rigid map by population) implies mean center. The simplified explanation of the next paragraph (shortest everage distance) implies geometric median. To round this out, the "Definitions section" at the end of the page (weighted average of coordinates) implies the median center method though it only mentions north and south. Considering the contradictions in this short document, it seems difficult to be certain of the method calculation, even from what should be one of the most reliable sources for such information. --Pikkusiili (talk) 12:39, 3 January 2014 (UTC)