> fpprec:n > > You need to write bigfloats like this: 1.0b0 In fact the bigfloats I would like to use are the solutions obtained with allroots(...). I want these numbers to be bigfloat and I don't know how ?