Open gjm112 opened 1 year ago
Can you fit a line through the n vs k optimal to come up with a simple formula for suggesting a value of k given a sample size?
k*_optimal = sqrt(2n) (see attachment to e-mail)
Can you fit a line through the n vs k optimal to come up with a simple formula for suggesting a value of k given a sample size?