Why does MATLAB returns a different value when we use a function?

Why do I get different ans values?

>> number = single(0.01);
>> floor( log10( abs( number ) ) )

ans =

      -2

>> func( number )

ans =

      -3

The func is described below.

function B = func( A )
    B = floor( log10( abs( A ) ) );
end

MATLAB should return -2 and does it on command window. But -3 is returned when we use function.

>> num2str( double(0.01), 17 )

ans =

0.01

>> num2str( single(0.01), 8 )

ans =

0.0099999998

Why does MATLAB correct it only by command window?

[Walter Roberson]

In R2016a it returns -2 . Which version are you using?

Answer 1

[Roger Stafford]

As you may be aware, in neither double precision nor single precision floating point format numbers such as you are using here, can the number .01 be exactly represented, since your computer uses binary numbers, not decimal numbers. However, double precision can come very much closer to it than single precision, and that accounts for the difference in results. It has nothing to do with whether a command window is used or a function performs the computation. It depends entirely on those numbers' formats.

[Yuhei Yabe]

Thank you for your answer.

> It depends entirely on those numbers' formats.

I think so, too. But, it returns -2.