Geographic Information Systems: An Introduction, 3rd Edition

14.9 Classification and Reclassification

To obtain a better view of the attribute data, they are often divided into classes. In the classification process, attributes are grouped according to limits set by the user. For example, three classes may be set for the attribute “year”: A = before 1970; B = 1971–1980; C = 1981–1990. Each object with a year attribute, such as a water supply pipe, is then assigned a new year-class attribute, A, B, or C. Plotting the classes in distinctive symbols or colors may reveal new patterns—showing, for instance, that water supply pipes in class B are those most subject to leakage.

Where the whole observation material is available, it is simple to divide the data into other classes, through reclassification. Reclassification can be desirable in order to see a new pattern and connections. In the preceding example, it can be possible to discover new patterns if one looks at a five-year period instead of ten-year period. Reclassification involves changing attribute values without altering geometries (Figure 14.13). Arithmetic and some statistical operations are used to assign new attribute values. In many ways, reclassification may be compared to changing colors on a map, in that the reclassified attributes may plot out in new colors. Reclassification may be used to isolate object types. For example, in a raster GIS, a “built-up area” characteristic may be isolated by assigning all other areas a value of zero (Figure 14.12). When the data are plotted ...