Configuring the Cause+ block

Properties

Name – The name of the specified Cause+ scenario to be loaded into the block.
Target – The selected target field. This is defined in the creation of the Cause+ scenario in the Continuous Troubleshooter.
Lower/Upper limits– The lower and upper limits defined by the High high and Low low limits specified in the Continuous Troubleshooter.
Trigger duration – The number of seconds which the cause fields are allowed to be above/below the upper/lower limits before the Cause+ scenario is triggered.
Calculation window– the number of seconds before the variable contribution of each field is calculated.
Messages– denotes the number of messages that were defined on each field during the creation of the Cause+ scenario in the Continuous Troubleshooter.
Cause ranking method- select to rank the causes according to either the:
- Hotelling T 2 value
This value explains how the values of the principal components are deviating, and what is causing the variation in these components. By selecting this ranking method, it is easier to identify which input field is having the greatest effect on variation in the principal components. The higher the T² value, the more distant the variable is from the model.
- Variable contributions calculated from the Principal Component Analysis.
These values explain how the values of the input fields - other than the principal components - are deviating, and what is causing the variation in these components. By selecting this ranking method, it is easier to identify which input field is having the greatest effect on variation in the other input fields in the process.
The number of top causes can be set by editing the box on the bottom left. The number is limited by the total number of required fields in the scenario.

Configuring the Cause+ block

Load/select the .CausePlus file containing your Cause+ scenario.
- This file was exported from the Continuous Troubleshooter.
- Highlight the required scenario, and click [Ok].
On selecting the scenario, the scenario name and Target field text boxes will be automatically populated.
Decide which method to use to rank the Cause fields. Read more about the different ranking methods in the Properties section above.
Map all the required fields in the Cause+ scenario, to the available double fields in the input port. Remember, the input port may contain only double type fields.

The input port must contain all fields used in constructing the loaded Cause+ scenario, and these input fields must be mapped to the required fields of the Cause+ scenario.
Select the number of top causes to be calculated.

The Cause+ block and data quality

There are multiple inputs to the Cause+ block, each of which can affect the outputs of the block in terms of data quality. The quality of the Cause+ block's output fields is determined by the following:

Whenever the Target field data quality is bad in a row, no calculations can occur and all block output fields are bad quality for that row. Note: If the Target field is outside its process limits with a good quality data value, and subsequent data values are of bad quality, the trigger duration will continue.
When any of the Cause fields are bad quality and a calculation window of 0 is used, (i.e. calculations are performed on each row of data), all block output fields will be bad quality.
When a calculation window is used, the block output fields will be bad quality only if any one of the cause fields had bad quality data over the entire calculation window.

Note: When a field in the dataset is bad quality, but that field is NOT used as any of the Cause fields nor the Target field, it does not impact the quality of the block.
If the total time window read is of a shorter duration than the calculation window, (possibly at the start of calculations) then calculations will not yet have had time to complete, and the output fields will be marked as bad quality.
If the total time window read is equal to the calculation window, then calculations will not yet have had time to complete, and the output fields will be marked as bad quality.
If the total time window read is greater than the calculation window, then calculations will have completed, and output cause fields will be calculated.