Open jeanbellynck opened 9 months ago
Can the clustering be "wider" in the extra dimensions?
Also, why not use System.Numerics.Vector
Also, why not use System.Numerics.Vector
?
Good point actually. I only did a quick search and did not find any vector class. I haven't closed this issue yet because I would like to replace my class with the vector class even if the performance gains will only be minor.
The program becomes very slow when running in high dimensions. I guess this is due to the Barnes hut algorithm. Every cube has 2^n subcubes stored so the memory might be increasing exponentially with the dimension.