Spatial Sampling Designs - Sampling Spatial Units for Agricultural Surveys - page 180

Fig. 7.7 Sample selected with LPM 1 ( on the left ) and LPM 2 ( on the right ) methods
> sam_lpm1 < - lpm1(p,X)
> length(sam_lpm1)
[1] 100
> spbalance(ds,rep(n/N,N),sam_lpm1)
[1] 0.09757576
> sam_lpm2 < - lpm2(p,X)
> length(sam_lpm2)
[1] 100
> spbalance(ds,rep(n/N,N),sam_lpm2)
[1] 0.1034343
> par(mar¼c(1,1,1,1),mfrow¼c(1,2),xaxs¼"i",yaxs¼"i")
> plot(framepop$xc,framepop$yc, axes¼F,cex¼0.5,pch¼19,
+ xlim¼c(0,1),ylim¼c(0,1))
> points(framepop$xc[sam_lpm1],framepop$yc[sam_lpm1],pch¼1,
+ cex¼2)
> box()
> plot(framepop$xc,framepop$yc, axes¼F,cex¼0.5,pch¼19,
+ xlim ¼ c(0,1),ylim ¼ c(0,1))
> points(framepop$xc[sam_lpm2],framepop$yc[sam_lpm2],pch ¼ 1,
+ cex ¼ 2)
> box()
The output (sam_lpm1, sam_lpm2) is again the vector of the codes of the
selected units. The SB index is lower than for SCPS, with a very small disadvantage
that LPM 2 is only a quicker and approximated version of LPM 1.