Performance Considerations in Data Windowing

There may sometimes be several ways of achieving the same result when obtaining a data point; however, whenever you retrieve a window of data, you should be aware of potential performance issues. The more data points that you retrieve, the slower your process will run.

For example, the following two processes both do the same thing: they compare the first and last value of a given time period.

Process 1

Process 1 is a visually simple representation of this process. It retrieves a 24-hour window of data, gets the first value of that window, and then compares it with the current value.

Process 2

Process 2 is more complicated. Rather than retrieving the entire 24-hour period of data, it only retrieves data for the minute before the start of that time period. It then takes the last value of that minute of data, and compares it with the current value.

Even though Process 2 appears to be more complicated than Process 1, it is in fact faster as it retrieves fewer data points.

Related: Sentinel Studio PrimersCreating a User Process, Data Functions, Data Set Functions, Usage Examples

Comments are closed