The fingerprint deviation tool helps you by identifying which tags are deviating from your chosen profile. This could be, for example, a golden batch profile or an ideal start-up behaviour.
When saving a fingerprint, you have the option to include relevant tags to the fingerprint that are not visible as a hull. Though hidden fingerprint tags are not evaluated in a fingerprint monitor or shown when visualizing a fingerprint, their evolution within the optimal profile is stored as a hidden hull by TrendMiner.
The fingerprint deviation tool will check both the visual and hidden hulls and list any deviations from the ideal behaviour for you automatically.
How to use the fingerprint deviation tool
- As a first step, you need to prepare your current view by visualizing the period you wish to compare with optimal behaviour.
- Open the diagnose menu.
- Choose the fingerprint deviations option.
- Select the fingerprint from the work organizer.
- In the fingerprint overview you will see a list of tags which are included in the fingerprint (either as visual or hidden). It is advised to select the option "Start of view + Zoom to fit" to add the fingerprint to the focus chart. This will make sure the complete fingerprint is visualized.
- Verify that the start of the fingerprint correctly aligns with the start of your period of interest. If this is not the case, go to the layer menu to adjust the position of the fingerprint.
- Fill out the necessary settings (see section below) and click "Diagnose" to see the results.
Note: The fingerprint needs to be visible from start to end for the diagnosis to work. A warning message will be shown if the fingerprint is not completely visible.
All tags that were saved within the fingerprint will be checked for any deviations and returned in the result list. You have the option to add tags from the result list to your chart which will also visualize a hull so you can inspect the deviation in better detail.
Fingerprint deviations settings
Saved fingerprint: This is where you select a previously saved fingerprint. Clicking this component will open the work organizer in a side panel pre-filtered on your saved fingerprints. Alternatively, you can add a saved fingerprint to your view before starting the diagnose flow. In this case, the visualized fingerprint will already be indicated.
The fingerprint needs to be completely visualized from start to end in your TrendHub view for the diagnosis to work.
Detection threshold: Changing the value affects the sensitivity when a deviation is detected. A high value means that the tags are not allowed to deviate much from the hulls while a lower value means that the tags are allowed to deviate more.