That was one of my first computer programs: compute a 'perfect' Yagi for a given frequency. Of course it will never be perfect because you will use tubes with a non-zero diameter and there will be something to carry the reflectors on and so on but it's amazing how close you can get if you cut and space precisely.