Clustering automatically groups points by numerical similarity, which can help you to quickly reveal distinct groups within your data. This can be useful for situations like identifying customer segments within sales data or categorizing IT tickets to identify common problems.

The underlying algorithm used in Virtualitics Explore is the k-means algorithm. Computed clusters can be saved as a new data feature, and for easy differentiation, are automatically mapped to the Color dimension in Mapping View. 

Important: Clustering works with numerical features but does not allow features where all the values are the same.

Using Clustering

1. Click the Clustering icon ( ) in the left-side toolbar to open the Clustering panel.

2. Input the number of clusters you'd like to display. By default, Number of Clusters is set to Auto. If you would like to specify an exact number of clusters to find, enter a number between 2 and 16.

   Note: When set to Auto, Clustering automatically determines the best number of clusters using the elbow method.

3. Select the features you would like to use to determine the clusters by choosing them from the Select Features drop-down or using the Input All button. 

4. (Optional) Remove unwanted features by hovering over the feature and clicking the Remove  button ( ).

   Tip: Removing sparse features (with many missing values) will produce more complete results. Click the red pound sign ( ) to remove sparse features.

5. Click Run.

   Important: You must have at least one data feature mapped to the X, Y, or Z plot dimension in order for clustering to display in Mapping View. One solution is to first run Smart Mapping, or create your own mapping in the Mapping Panel.

Next Steps

Once you've run Clustering, you'll see that computed clusters are automatically mapped to the Color dimension. A new data feature is also generated called (X) Cluster Result, where X corresponds to the number of times you have run Clustering. You can simply right click the newly-created feature and select Add to Features, which will add this feature to the Features and Filters Panel, allowing you use this feature in future mappings.

Based on your mapping settings in Mapping View, you can easily view the clusters on your plot. Additionally, you can use the Insights tool on your newly created Cluster Result data feature to see additional information about that data.