buildFrameSpatial: Builds the "sampling frame" dataframe from a dataset...

In barcaroli/SamplingStrata: Optimal Stratification of Sampling Frames for Multipurpose Sampling Surveys

Builds the "sampling frame" dataframe from a dataset containing information all the units in the population of reference including spatial

Description

This function allows to build the information regarding the sampling frame of the population of reference. Mandatory variables are: (i) the name of the dataset containing the sampling frame of the population of reference (ii) an identifier (Id) (iii) a set of auxiliary variables 'X' (iv) a set of target variables 'Y' (v) a set of prediction errors variables 'variance' (vi) longitude (vii) latitude (viii) the indicator of the domain to which the unit belongs

Usage

buildFrameSpatial(df, id, X, Y, variance, lon, lat, domainvalue)

Arguments

df	This is the name of the dataframe containing the information on all the units in population of reference.
id	This is the name of the identifier in the sampling frame.
X	A character vector containing the names of the auxiliary variables in the frame dataset
Y	A character vector containing the names of the target variables in the frame dataset
variance	A character vector containing the names of the prediction error variables in the frame dataset
lon	Longitude of the unit
lat	Latitude of the unit
domainvalue	The name of the variable in the frame dataset that contains the indication of the domains to which the units belong.

Value

A dataframe

Author(s)

Giulio Barcaroli

Examples

## Not run: 
library(sp)
library(gstat)
library(automap)
library(SamplingStrata)
data("meuse")
data("meuse.grid")
meuse.grid$id <- c(1:nrow(meuse.grid))
coordinates(meuse)<-c("x","y")
coordinates(meuse.grid)<-c("x","y")
#################
# kriging
#################
v <- variogram(lead ~ dist + soil, data=meuse)
fit.vgm <- autofitVariogram(lead ~ elev + soil, meuse, model = "Exp")
plot(v, fit.vgm$var_model)
fit.vgm$var_model
g <- NULL
g <- gstat(g, "Pb", lead ~ dist + soil, meuse)
g
vm <- variogram(g)
vm.fit <- fit.lmc(vm, g, vgm(psill=fit.vgm$var_model$psill[2], 
                             model="Exp", range=fit.vgm$var_model$range[2], 
                             nugget=fit.vgm$var_model$psill[1]))
# Prediction on the whole grid
preds <- predict(vm.fit, meuse.grid)
names(preds)
# [1] "Pb.pred" "Pb.var"
preds$Pb.pred <- ifelse(preds$Pb.pred < 0,0,preds$Pb.pred)
df <- NULL
df$id <- meuse.grid$id
df$Pb.pred <- preds@data$Pb.pred
df$Pb.var <- preds@data$Pb.var
df$lon <- meuse.grid$x
df$lat <- meuse.grid$y
df$dom1 <- 1
df <- as.data.frame(df)
frame <- buildFrameSpatial(df=df,
                      id="id",
                      X=c("Pb.pred"),
                      Y=c("Pb.pred"),
                      variance=c("Pb.var"),
                      lon="lon",
                      lat="lat",
                      domainvalue = "dom1")
head(frame)

## End(Not run)

barcaroli/SamplingStrata documentation built on Oct. 13, 2023, 8:56 a.m.