IDSL-VAL-008: Exact match on a real number¶
| Default severity | Kind | Scope | Quick fix |
|---|---|---|---|
| info | static | specification | no |
The value is an exact real number on a floating-point data type. IDS 1.0 compares reals with a small relative tolerance (1e-6, see the buildingSMART tolerance test cases), which absorbs rounding but not unit conversion or modelling tolerance: 2.4 does not match 2.4004. A range states the tolerance explicitly. Tolerance handling is under discussion for IDS 1.1 (IDS issue #418).
Example¶
References¶
Suppressing¶
Add a suppression with a reason for IDSL-VAL-008 (or IDSL-VAL-*) to the node, its facet, its specification or the document in the Studio sidecar (meta.suppressions).