When mapping large point datasets, it's common to see your points overlap one another. But what happens when most of the points within the map are covering one another? The map starts to look like noise, rather than an informative narrative about the data. 

This is a missed opportunity. 

ArcGIS offers ways to get around this fundamental spatial issue, such as aggregation methods, heat maps, and statistical analysis. These methods help to lasso the data, but many times require pre-processing. 

New to ArcGIS Online, you can now cluster point features on-the-fly to easily visualize your large point Feature Layers. 

Clustering allows you, as the map maker, to explore and visualize patterns that would have otherwise been hidden. Quickly see your points aggregated into smaller groupings of points. This provides a better understanding of how many points exist within an area.

As you zoom in and out, the clustering will change based on your current extent. Clusters are proportionally sized by the count of features within each cluster. This means that smaller symbols have fewer points clustered within them, and larger symbols have many features clustered within them.

Let's see an example in action:

This video also provides a quick tutorial showing how to get started with clustering your points!

Clustering allows you to visualize the quantity of points within smaller groupings. But clustering also allows you to maintain your existing cartography.

 Clustering Categorical Attributes 

Below, we see a map of police incidents in the Philadelphia area. This is mapped using the Types (Unique Symbols) option. The map shows each color as a distinct type of incident that has occurred. Unfortunately, there are so many incidents in this area, it is hard to make sense of this map, and no immediate patterns of crime are apparent.

When clustering is enabled on this Categorical map, the predominant incident type will be represented by the feature. The size still represents how many features were clustered. Now, we can see that there are many thefts in the downtown area.

When clustering a categorical map, the default clustering popup will still show the count of features in the cluster, but will also provide the predominant value of the features being clustered. 

Clustering Numeric Attributes

Below, we see a map which shows transit stops in the Los Angeles area. The brighter yellow stops have more hourly trips during the rush hour 7-9 am time slot. Points in blue have fewer stops per hour. Again, there are many overlapping transit stops, so it is hard to decipher a pattern (especially in the downtown area).

When this layer is clustered, the colors retain their meaning, with the spectrum of yellow to blue. But now, we get a better understanding of the amount of transit stops AND the amount of buses that visit those stops during rush hour. The value dictating the color represents an average of the stops being clustered.

In this case, the default clustering popup portrays the average value of the transit stops within each cluster.

*When using clustering, you can set your cartography before or after turning on clustering. Whatever works best in your everyday workflows! These examples showcase what happens when your cartography already exists.

The new clustering option in ArcGIS Online allows you to quickly and easily discover new things about your point Feature Layers. Maintain your cartography, arrange the amount of clustering, and watch the patterns unfold! 

For more information visit the ArcGIS Online Help Pages.