E is the result of the builtin function std::exp(1).
The precision of this value is therefore system-dependent.