The purpose of this work is to present the idiosyncrasies of floating-point arithmetic, hardware and software implementation constraints, the origin of computational errors, the numerical method instability problem and finally an interactive language for teaching theory of errors.
