Sets the objects error tolerance value, epsilon. Default value is . Usage: fix_gilbert( float_value ) where float_value is the desired error tolerance value.