Data Metrics Data Location Statistics
There are five groups of data metrics: Z Order Statistics, Z Moment Statistics, Other Z Statistics, Data Location Statistics, and Terrain Statistics. The Data Location Statistics calculate information about the points around each of the nodes.
The separation distances between the current grid node and each of the selected data are used in the computation of the data location statistics. In the following discussion, let the location of the current grid node be represented as 
. The list of separation distances are defined as
|
|
The
are sorted in ascending order and indexed as
|
|
|
Data Metric |
Definition |
Equation |
|
Count |
The Count data metric generates an output grid for which each nodal value is the number of data selected by the specified sector search centered at that node. Under most circumstances, the best interpolation results occur when the Count is approximately homogeneous throughout the area of interest. |
|
|
Approximate Density |
The Approximate Density data metric generates an output grid for which each nodal value is the number of data selected by the specified sector search centered at that node, divided by |
|
|
Distance to Nearest |
each nodal value is the distance to the nearest datum selected by the specified sector search centered at that node |
|
|
Distance to Farthest |
each nodal value is the distance to the farthest datum selected by the specified sector search centered at that node |
|
|
Median Distance |
each nodal value is the median separation distance between the node and all of the data selected by the specified sector search centered at that node |
|
|
Average Distance |
each nodal value is the average separation distance between the node and all of the data selected by the specified sector search centered at that node |
|
|
Offset Distance |
each nodal value is the distance between the node and the centroid of all of the data selected by the specified sector search centered at that node |
|
, where 
is the distance from the node to the farthest selected datum. The area over which the density is computed is bounded by the distance to the farthest selected datum. As such, when the data are relatively sparse, the computed density is slightly over-stated.
