To specify an exceeded standard deviations component
*
These steps describe how to build an exceeded standard deviations (ESD) component.
1.
2.
Set the Number of Standard Deviations.
The Number of Standard Deviations value sets the number of standard deviations above (or below) the historical value for whatever time period you are looking at. The value can be a constant value or it can be custom to the customer based on a lookup value from a datasource or on a calculated math value.
The value can be any of the following.
*
*
*
*
*
3.
In the Data Source area, select the data source by type and name, and identify the field on which to do the trending.
4.
In the Spike Value area, define the spike period and function.
For example, you might define the spike value to be the average of two rolling days.
5.
Define the Historical Value for Comparison, which is the value to which the spike value will be compared.
For example, you might define it to be the average over three weeks.
6.
You can optionally set Subgrouping within trend periods.
If you want to make the data function you are using act on subgroupings of values rather than on individual values in the container, then select the checkbox Subgrouping within trend periods and set the Subgroup Period and Function.
7.
You can optionally set Advanced Configuration settings for the ESD trend. Advanced configuration settings can help to ensure that data has been tracked long enough. They can also ensure that you have enough data points, because the ESD does require a reasonable number of data points in the historical period for the concept of a standard deviation to be significant.
8.
Click Set Parameters to save the details settings.
9.
On the Properties tab, set the name and description of the ESD trend.
10.
Click Set Parameters to set the properties settings, click Save to save the component, then click Close to close the editor window.