In the Latin hypercube sampling, the distribution is obtained using the results of goodness of fit tests.
#Progressive latin hypercube sampling series
Basically, I take a random permutation of the grid cells first, then sample within the cell on the right portion of a uniform, then finally transform that margin to the desired distribution (normal, lognormal, etc). Unlike Latin hypercube sampling, PLHS generates a series of smaller sub-sets (slices) such that (1) the first slice is Latin hypercube, (2) the progressive union of slices remains Latin hypercube. in positioning bolt-holes in the loading direction of connections, the statistical technique of sampling known as Latin hypercube sampling has been adopted in this study. Take a look at the lhs package in R which I maintain: here and here. I agree that with only two variables, it is not necessary to "shuffle" them both, however, with multiple variables, the cells are normally determined by a random permutation of the margins. Instead of shuffling at the end, you can also randomly permute the cells in the beginning.
0 kommentar(er)
