To perform a radial analysis, you must first have assigned lat/long coordinates (geocodes) to your point files. You can geocode your files with the GeoStan library, or the Geocoder or Centrus Desktop applications.
In a radial analysis, there is one set of points representing the items to be searched—a set of points against which other points are tested. In many applications, this set of points is store or office locations. This file is referred to as the "site" file. Site files can also be created from the points in a Spatial+ object file.
There is also another set of points, which are tested against the site file. Commonly, this is a customer file, or new prospect file, etc. This file must also have lat/lon coordinates. This file is referred to as the "target" file.
In order to optimize a radial search, the site file must be indexed to form a search tree. The Spatial+ library provides functions to do this rapidly. The search tree is built in memory, and is not stored to disk. Each point occupies 20 bytes, in addition to the data stored with it.
The radial functions were developed to help spatially relate points within a general area. The distances that are calculated using radial functions are very accurate for points within 50 or 100 miles of each other. As the north-south distance becomes greater, slight errors in calculations can occur. The following examples are for points that are oriented directly north or south of the base point, which would have the largest error for the distance given.
| North-South Distance | Maximum Error |
| 50 miles | 0.25% (.1 miles) |
| 200 miles | 0.9% (1.8 miles) |
| 400 miles | 1.25% (5 miles) |
| 1,000 miles | 1.9% (19 miles) |
The following functions are used by radial analyses:
- pipRadInit
- pipLoadPoints
- pipRadInsertPoint
- pipRadFindFirstInRect
- pipRadFindFirstInRadius
- pipRadFindFirstNearest
- pipRadFindNext
- pipDistance
- pipRadTerm
For complete information on how to perform a radial analysis and sample code, refer to the pipRad_ functions.