How high-precision calculations can stimulate mathematical research (Q1195358)

The author discusses three cases in the area of rational approximation theory where theoretical studies and high-precision calculations stimulated each other. As the author was personally involved in these developments, the discussion is quite lively. The first case is the approximation of \(e^{-x}\) uniformly in \([0,\infty[\) by rational functions of prescribed degree. The second case (the Bernstein conjecture) deals with \(| x|\) in \([-1,+1]\). In the third case, \(| x|\) is approximated in \([-1,+1]\) by rational functions. In each situation, high-precision calculations either led to the formulation of conjectures or allowed to disprove them. The author also presents an unsolved conjecture based on numerical results.