nan()
Constant number representing Not A Number, generally used for comparison to the result of mathematical operations with invalid inputs. This is equivalent to a decimal (floating point) null, but nan() is provided for more explicit usage in mathematical expressions.
Decimal
nan(100)
returns null