Code: Select all
positions = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
engine1_bm = [a, b, c, d, e, f, g, h, i, j]
engine2_bm = [a, x, c, d, e, v, t, r, i, s]
engine3_bm ...
...
Similarity of bm between engine1 and engine2 is:
Similarity = a, c, d, e, i
Similarity = 5/10 = 50%
Then compute similarity between engine1 and engine3, engine1 and engine 4 and so on.
So it would look like this as in sim.
Code: Select all
------ Stockfish 10 (time: 20 ms scale: 1.0) ------
68.67 Stockfish 10 r1 (time: 20 ms scale: 1.0)
38.06 Senpai 2.0 (time: 200 ms scale: 1.0)
37.63 Demolito 2018-10-29 (time: 200 ms scale: 1.0)
37.11 RofChade Version 2.0 (time: 200 ms scale: 1.0)
37.10 Texel 1.07 (time: 200 ms scale: 1.0)
36.94 Gull 3 x64 (time: 200 ms scale: 1.0)
36.70 Arasan 21.1 (time: 200 ms scale: 1.0)
35.28 Wasp 3.50 (time: 200 ms scale: 1.0)
35.26 Booot 6.3.1 (time: 200 ms scale: 1.0)
34.58 Atlas 3.91 (time: 200 ms scale: 1.0)
34.38 Hannibal 1.7 x64 (time: 200 ms scale: 1.0)
33.41 Andscacs 0.95 (time: 200 ms scale: 1.0)
31.28 Stockfish 9 (time: 20 ms scale: 1.0)
31.03 The Baron 3.44 (time: 200 ms scale: 1.0)
30.18 Lc0 v0.21.1 w11258-120x9 (time: 5000 ms scale: 1.0)
29.93 Pedone 1.9 (time: 200 ms scale: 1.0)
29.15 Ethereal 11.25 (time: 200 ms scale: 1.0)
26.40 Vajolet2 2.6.1 (time: 200 ms scale: 1.0)
26.26 Fizbo 2 (time: 200 ms scale: 1.0)
25.69 Laser 1.7 (time: 200 ms scale: 1.0)
25.49 Cheng 4.39 (time: 200 ms scale: 1.0)
24.87 Deuterium 2019.2.37.37 (time: 200 ms scale: 1.0)
24.63 Xiphos 0.5 (time: 200 ms scale: 1.0)
20.45 Nemorino 5.00 (time: 200 ms scale: 1.0)
19.86 Amoeba 2.8 (time: 200 ms scale: 1.0)
19.26 Schooner 2.0.34 (time: 200 ms scale: 1.0)
18.06 SmarThink 1.98 (time: 200 ms scale: 1.0)
Currently I am using average or mean. I will use this data to construct a dendrogram.