Table 1.
A list of segregation measures currently implemented in the seg package.
Table 2.
Supported input and output classes for the implemented functions.
Figure 1.
Computational flow of the spseg() function.
It calls a series of subfunctions to calculate the spatial segregation measures. In this diagram, the curved-rectangles represent R functions and processes, the parallelograms refer to R objects, and the diamonds indicate the user options. Among the rectangles, only the shaded ones are user-level functions.
Figure 2.
Available methods for SegSpatial and SegLocal.
SegSpatial is a S4 class that stores results from the spseg() function. It inherits from another S4 class SegLocal.
Figure 3.
Available methods for SegDecomp.
SegDecomp is a custom defined S4 class, containing the measured segregation from deseg().
Table 3.
A list of hypothetical segregation patterns adopted in this paper and their original citation.
Table 4.
,
,
and
for the hypothetical segregation patterns listed in Table 3.
Figure 4.
A sample pattern of segregation.
The white and black cells are where the minority population comprises 0% and 100% of the local population, respectively. The numbers inside of the cells indicate the cell ID, and the letters denote the edges.
Table 5.
,
and the surface-based spatial segregation measures for the hypothetical segregation patterns listed in Table 3.
Figure 5.
Computation time of the implemented functions for different input size.
As the number of spatial units increases, the computation time also increases for all functions but at a different rate. The functions tested here are: dissim() (A), conprof() (B), isp() (C), spseg() (D), and deseg() (E).
Figure 6.
Relationship between the computation time and the number of measurement points for spseg().