seriesTransform-SeriesInterval

When creating a transform graph, it is important that the data conform to exact intervals to allow data from different history sources to properly compare and combine into larger data sets. To do this, a process of “quantization” is used to fit the timestamp value of a record to an exact interval.

In the case of histories, the timestamp of each history includes not only the hour, minute, and second that the history was created, but also the millisecond value. When attempting to align different histories for comparison, this difference in milliseconds prevents an exact comparison. The History Source window can set intervals to predefined increments using the Interval option list.

Figure 44.   Interval and Date Range fields
Image

Setting the Interval to Automatic selects the current interval for history collection. For example, if the history is already set to collect at 15 minute intervals, the framework also sets the Automatic quantization interval of the history source node to 15 minutes.

 NOTE: All values are quantized based on a starting time of 0:0:0.000. This means that an interval of 15 minutes quantizes a record whose timestamp reflects the value of 11:13:52.093 to 11:00:00.000. 

To use only a segment of a History rather than a point value’s entire history, use the Date Range property and select the starting and ending date and times for the history data. The framework includes in the data set at the time the graph is resolved all history records between these date values, including those histories that fall exactly on those date values.