TH NextGen - Layer comparison

The layer comparison functionality allows you to quickly compare tags or attributes of interest for different periods of concern, in a numerical and visual way. You will need to create and overlay different layers or periods (either manually or via searches) in order to perform a numerical layer comparison.

In practice the layer comparison tool can be used to:

  1. Determine root causes by overlaying good and bad periods, numerically compare layers and browse for differences.
  2. Gain process insights by comparing different regimes of operation.
  3. Quickly identify periods with the highest throughput efficiency.

This article describes how to use the layer comparison table and the different statistical parameters available for layer comparison.

Using the Compare Layer table

The Compare layer table option can be found on the left side of the view bar next to the Statistics table.

Once a tag or attribute is added and is active it can be visualized in the compare table. The table shows, by default, the average value of all active tags. The compare table can be opened when there is only one layer, but the main value of the statistics table is to easily compare different layers.

When additional sublayers are present, statistical values are calculated for all layers. The sublayers are compared to the base layer, and the comparison is done by providing a relative difference for all sublayers compared to the base layer, i.e.:

difference = (Value_sublayer – Value_baselayer) / |Value_baselayer|

Negative differences are indicated in blue, while positive differences are indicated in orange.

Compare.1.0.png

You will also notice that the compare layer table is automatically updating the selected parameter and comparison/contrast is shown in the table when a change to the focus chart is detected.

Table options

The values in the table are automatically updated when making any time related change on the focus chart.

Within the table you can perform a few actions:

  • Sorting columns
  • Adjusting the size of the columns
  • Hiding tags

Each column can be sorted, which is easily done by clicking the column header. Clicking on a header highlights the applied sorting used, for example, an up/down arrow icon (ascending / descending) or no arrow (sorting not active). One column can be sorted at a time, including the Tags & Attributes column. If no sorting is applied, the order of the tags corresponds to the order in the active tag list.

You can alter the size of the widths of the columns, and attributes can be made visible or hidden using the eye icons next to the data reference names.

Note: Hidden tags and attributes are also displayed in the compare layer table.

Layer options

Multiple options which are accessible in the layer menu can be performed directly in the compare table as well.  The following actions are available:

  • Hide / Unhide layer
  • Rename layer (option only for the base layer)
  • Set as base layer
  • Remove layer

Compare.1.3.png

Compare.1.4.png

Compare layer table settings

The compare layer table has a couple of settings that can be updated. To open the compare layer table settings panel, click on the blue settings button on the top right of the table.

Compare.1.1.png

On the "Edit table to compare layers" panel, you can select the statistical parameters of interest you wish to see in the table.

 

accuracy_comparison.png

Key Parameters

By default, the "Average (AVG)" statistic is enabled for the compare layer table. This statistic can be switched to any of the other key parameters, including similarity parameters.

Most statistical parameters for comparing layers are available in the statistics table. More information about key statistics can be found on the statistics table article.

Similarity Parameters

These parameters are specifically used to compare active layer values with each other. Below you can see some more details about these extra parameters.

Statistical similarity
The statistical similarity is a calculation which compares the distribution of the data between layers. The value is returned as a percentage. The higher the value, the more likely both layers have the same distribution.

This similarity score is based on the Kolmogorov-Smirnov test statistic.

Evolution similarity
This parameter measures the evolutional similarity between the layers in function of the base layer, which results in a percentage value. The higher the value, the more similar the evolution throughout the layer.

Correlation
This parameter is a measure of linear correlation between the layer in question and the base layer. The parameter corresponds with the Pearson correlation coefficient.

Precision

This option allows you to fine-tune the number of digits after the decimal point for the values in the table. By default, this value is set to three with a maximum value of eight.

Units of measurements

Contrary to the statistics table, the units of measurements are not listed under the parameters but is another option. Enabling this checkbox, will show the units of measurement for the used tags or attributes (if available).

Enjoying TrendMiner? Share your experience to help the community.
Rate TrendMiner