Pitney Bowes MapInfo Vertical Mapper - Manual (Page 183)

Chapter 10: Aggregating Data
User Guide
181
new aggregated point. The process ends by performing the aggregation calculations on the selected
points as specified in the Point Aggregation dialog box, and the results are attributed to the new
aggregated point.
Once the first point is processed, the procedure sweeps from left to right and top to bottom across
the study area, selecting and aggregating unflagged points. It is important to note that not every data
point will be aggregated on the first pass through the data set. Normally a second pass is required to
aggregate those points missed on the first pass. The results are shown in the next figure.
Example of circle aggregation using the Forward Stepping Aggregation technique. The
shaded points are the original data points. The crosses represent the newly
aggregated points, and the circles represent the aggregation region. The original
points have been coded to show to which aggregation region they belong.
In the figure above, you may notice that inappropriate aggregation decisions have been made in
certain locations of the point file as well as the degree of overlap of the aggregation regions. In the
upper left corner of the diagram there are two examples of an inappropriate aggregation, marked by
the letters A and B. In both cases you would aggregate these points differently if you performed this
process manually. The reason these points are aggregated this way has to do with the two
aggregation passes this technique performs; the second pass aggregates the remaining unflagged
points, resulting in a large degree of overlap of the aggregation regions. Some of the aggregation
regions in the above diagram have been numbered to show the aggregation process order. The
letters show which points were aggregated on the second pass.
Cluster Density Aggregation
The Cluster Density Aggregation technique is typically used when a visual clustering effect occurs in
the dispersion of the data points. For example, demographic data representing rural areas may often
exhibit a ‘shotgun’ pattern (see the next figure). Data from each small community is considerably
more densely distributed than in the surrounding countryside. Cluster density does, however,
process large data sets more slowly than the other aggregation techniques.