Many applications require the discovery of patterns in data representing a sequence of events; examples include quality control and fault diagnosis and prevention in industrial and mechanical processes. Data in these applications typically takes the form of logs; that is time-stamped sets of measurements that form a sequence. The measurements may be very simple, even a single variable, but the patterns are found in how these measurements vary over time. Modeler includes a variety of features for processing sequential data of this sort. This recipe illustrates some of these sequence processing operations and how they are used to build up a set of variables describing the changes in measurement over time.
This recipe requires no datafile because the example data is generated by a user input source node and other operations inside a source supernode. The stream file required is Sequence_Processing.str
.