Hello

1) I'm trying to distribute m points over an n-cube in such a way that they cover as best as possible the hypercube, ie:
. the average (really integral) distance of all points in the hypercube to their respective nearest point in the m-cloud is minimized

2) If j points of the m-cloud are already fixed, how to choose the remaining m-j points to satisfy the above condition

Can this be found in "Computational geometry"?

regards,