# systematic: returns a systematic exploration of a vector In rbsb: set of ancillary functions

## Description

returns a vector of indices of `x` such that this is explored in systematic way by `nb` draws, spaced of `kk` according to its order.
Even is `ord` is not "random", a call to `sample` is done so it is advised to call first `set.seed`.

## Usage

 `1` ```systematic(x, nb, kk, ord="random") ```

## Arguments

 `x` The vector to systematically sample. `nb` desired number of draws. `kk` frequency of successive draws. Something giving quasi-uniform draw is is ```min(length(x)-1,length(x)% /% (nb% % length(x)))``` ???notice that the direct application of this formula failed for rounding phenomenum???. `ord` indicates which type of ordering must be done on the output: `randor` for a circular random permutation; `random` for a complete random permutation; `order` for a sorting with `sort`; `such` for nothing (in that case, it depends on the algorithm which can be

## Details

When `nb*kk` is less than `length(x)` then the sequence is 1, kk+1, 2*kk+1,..., (nb-1)*kk+1 taking into account the ordered order of values of `x`. If not either look at the code or experiment the function.

## Value

A vector of length `nb` of indices of `x`.

## Examples

 ```1 2 3 4 5 6 7 8 9``` ``` rbsb3k("reset"); # only necessary for R checking systematic(1:20, 5, 1); systematic(1:20, 5, 3); systematic(1:20, 15, 3); systematic(1:20, 15, 3, "order"); systematic(1:20, 15, 3, "random"); sort(systematic(1:20, 55, 3)); uu <- c(2*(1:10), 1+2*(0:9)); uu[systematic(uu, 20, 3)]; ```

rbsb documentation built on May 31, 2017, 1:45 a.m.