When we perform comparisons for pass/fail judgment, tolerances define how close the comparison must be for the judgment to be considered a pass.

We define the expected behavior of a system by specifying output Signal patterns.  If the Actual Output Signal patterns match their Expected patterns, then a TestCase is considered to have passed.  The question arises as to how close the Actual Signal pattern has to match the Expected Signal pattern.  We need to specify some kind of tolerance level to apply when we perform the comparison. There are 3 kinds of tolerances: Time, Value, and Count. They can be assigned to:

The time that something occurs

The value that it has at a point in time

The number of changes (or transitions)

The time between transitions (for Periodic Signals)

Which tolerances are applicable and what the tolerances mean in terms of the Pass/Fail Judgment depends whether the Signal is Numeric, Message, or Event, and for Numeric Signals, whether it is continuous or discrete.  Normally, discrete Signals do not  have a value tolerance. See Signal Properties.

Tolerance settings are visible at the bottom of the Transition List and on the TestCase page of Regression Reports.

hmtoggle_plus1Positive and Negative Tolerances
hmtoggle_plus1Tolerance Inheritance
hmtoggle_plus1Tolerance Bands


Related Topics:

Custom Tolerances

Pass/Fail Judgment Overview

Signal Properties

DataBlock Overview

TestCase Properties

TestCase Properties—Absolute Time Tolerances

TestCase Properties—Relative Time Criteria

Pattern Matching

Relative Time Criteria

Periodic Signals