Using a Collection Graph

Collections show the distribution of values for one numeric field relative to the values of another, rather than the occurrence of values for a single field. Histograms operate similarly to collections graphs. Histograms show the distribution of values in a numeric field whose values range along the x axis.

Once you have created a graph, you can examine the results and define bands to split values along the x axis or define regions. You can also mark elements within the graph. See the topic Exploring Graphs for more information.

You can use options on the Generate menu to create Balance, Select, or Derive nodes using the data in the graph or more specifically within bands, regions, or marked elements. This type of graph is frequently used before manipulation nodes to explore the data and correct any imbalances by generating a Balance node from the graph to use in the stream. You can also generate a Derive Flag node to add a field showing which band each record falls into or a Select node to select all records within a particular set or range of values. Such operations help you to focus on a particular subset of data for further exploration. See the topic Generating Nodes from Graphs for more information.

Figure 1. 3-D collection graph showing sum of Na_to_K over Age for both high and normal cholesterol levels
3-D collection graph showing sum of Na_to_K over Age for both high and normal cholesterol levels
Figure 2. Collection graph without z axis displayed but with Cholesterol as color overlay
Collection graph without z axis displayed but with Cholesterol as color overlay