Skip to main content

/

Docs

Nightly

v24.6

Log in

Mojo function

assert_almost_equal

assert_almost_equal[type: DType, size: Int](lhs: SIMD[type, size], rhs: SIMD[type, size], msg: String = String(""), *, atol: SIMD[type, 1] = SIMD(#kgen.float_literal<1|100000000>), rtol: SIMD[type, 1] = SIMD(#kgen.float_literal<1|100000>), equal_nan: Bool = False, location: Optional[_SourceLocation] = Optional(None))

Asserts that the input values are equal up to a tolerance. If it is not then an Error is raised.

When the type is boolean or integral, then equality is checked. When the type is floating-point, then this checks if the two input values are numerically the close using the abs(lhs−rhs)<=max(rtol∗max(abs(lhs),abs(rhs)),atol)abs(lhs - rhs) <= max(rtol * max(abs(lhs), abs(rhs)), atol) formula.

Constraints:

The type must be boolean, integral, or floating-point.

Parameters:

  • ​type (DType): The dtype of the left- and right-hand-side SIMD vectors.
  • ​size (Int): The width of the left- and right-hand-side SIMD vectors.

Args:

  • ​lhs (SIMD[type, size]): The lhs of the equality.
  • ​rhs (SIMD[type, size]): The rhs of the equality.
  • ​msg (String): The message to print.
  • ​atol (SIMD[type, 1]): The absolute tolerance.
  • ​rtol (SIMD[type, 1]): The relative tolerance.
  • ​equal_nan (Bool): Whether to treat nans as equal.
  • ​location (Optional[_SourceLocation]): The location of the error (default to the __call_location).

Raises:

An Error with the provided message if assert fails and None otherwise.