Thrown if the passed row collection
has a value that cannot be cast to a single-precision number.
Remarks
Each row in the resulting matrix is initialized with the values in a row
in the given row collection. All values must be able to be converted to
a float through a cast.