Modeler: 14/Grading

From CaseTalk Wiki
Jump to: navigation, search

Grading Preferences

The Grading tab in Tools > Preferences lets you configure how your model is graded. For background on what model grading is and how scoring works, see Model Grading.

Calculation Method

The radio group at the top selects the calculation method:

  • Average - mean of all scores. Every element counts equally.
  • Weighted - violations weigh more. Elements flagged by heavier rules have greater influence.
  • Worst-N - average of worst 10%. Focuses on the most problematic elements.
  • Minimum - worst element score. One bad element determines the entire grade.
  • Penalty - subtract from 100. Each violation directly reduces the score, making issues clearly visible.
  • None - grading disabled.

File:GradingTab.png

Rule Weights

The grid below shows all 25 well-formedness rules with:

Column Description
Checkbox Enable or disable a rule for grading. Click to toggle.
Rule The well-formedness rule description (read-only).
Severity Error (red), Warning (orange), or Hint (gray). Indicates the severity level.
Weight How much this rule affects the score. Higher weight means bigger impact. Editable.

Default weights:

Severity Default Weight
Error 15
Warning 8
Hint 2

Reset

Click Reset Defaults to restore all rules to enabled with their default weights, and set the method to Average.


Configuration
Dialogs Environment · Project Options · Tools Options · GLR Options · Custom Attributes · Grading
Reference Customization · Bookmarks and URIs · Multi Lingual · MCP Client