Skip to main content

Confusion matrix

Confusion matrix evaluates a model by comparing its predictions with actual results in a tabular format. It displays a table where each cell counts rows for a specific combination of predicted and actual values, while also showing prediction errors and overall model accuracy.

To add a confusion matrix, click the Add Viewer icon on the menu ribbon and select Confusion Matrix.

developers

To add the viewer programmatically from the console, use:
grok.shell.tv.addViewer('Confusion Matrix');

Confusion Matrix

Properties

PropertyTypeDescription
General
X Column NamestringColumn to be put on the X axis
Y Column NamestringColumn to be put on the Y axis
Row SourcestringDetermines the rows shown on the plot.
Allow Dynamic Menusboolean
Show Context MenubooleanProperties common for all viewers todo: use code generation
Titlestring
DescriptionstringViewer description that gets shown at the Descriptor Position. Markup is supported.
HelpstringHelp to be shown when user clicks on the ''?'' icon on top. Could either be in markdown, or a URL (starting with ''/'' or ''http'').
Description Positionflexposition
Description Visibility Modevisibilitymode
Data
FilterstringFormula that filters out rows to show. Examples: ${AGE} > 20 or ${WEIGHT / 2)} > 100, ${SEVERITY} == ''Medium'', ${RACE}.endsWith(''sian'')
Tablestring
Style
Controls FontstringViewer controls elements font.
Description
Show Titleboolean