[_0ffh]

Hi, in an unrelated nitpick: The relative error should be calculated by dividing the error by the true value, not by it's approximation. Still, very nice writeup!

[djk447]

Thanks! messed up the formula but had it right in the text :( Fixed now.