Number of observations in the lowest level (ie level 1, the fastest
code)
levels
Number of levels
prob
Probability of choosing an observation at level
\(n+1\) given that there is one at the same place at level \(n\)
Author
Robin K. S. Hankin (subsets.fun()); Peter Dalgaard (via R-help)
References
M. C. Kennedy and A. O'Hagan 2000. “Predicting the output from
a complex computer code when fast approximations are available”
Biometrika, 87(1): pp1-13