Data Classification Methods
Advanced data classification methods for both vector and raster layers were introduced in Developer Kernel v. 11.35 and desktop GIS Editor v. 5.18.
Classification refers to the process of grouping spatially located observations into data ranges, called classes. Classification is performed based on a specified attribute field, classification method, and number of classes. Data classification is primarily used to produce choropleth (thematic) maps, but also for analyzing datasets and displaying continuous field data.
The DK includes an impressive list of implemented classification methods:
- DefinedInterval
- EqualInterval
- GeometricalInterval
- NaturalBreaks
- Quantile
- StandardDeviation
and some less common methods that are also useful:
- Kmeans
- KmeansSpatial
- Quartile
A description of each method is found at: TGIS_ClassificationMethod.
The entire data classification process can be done at a low-level, using classes:
- TGIS_ClassificationVector
- TGIS_ClassificationPixel
This functionality is implemented as a ready-to-use feature in the DK and Editor's Layer Properties control. The control's Rendering Wizard dialog box (and the Grid Wizard contained within it) now present two data classification options:
- Simple, using earlier classification methods based on unique or continuous values, and
- Advanced, using the new classification methods discussed in this blog.
The wizard offers an easy way to experiment with and
learn about the new classification methods.
The overall functionality is documented at: TGIS_ClassificationAbstract class.
Country populations classified using the NaturalBreaks method and five classes.
A grid layer classified using the DefinedInterval method with Interval set to 250m.