## ----covtype_readdata----------------------------------------------------
covtype <- read.csv(file="Data/CoverageType/covtype.data", header=FALSE, sep=",", stringsAsFactors=FALSE)
head(covtype,n = 2)


## ----covtype_name--------------------------------------------------------
# "Fixed" attributes' names
names <- c("Elevation","Aspect","Slope","HorDistToHydro","VertDistToHydro",
           "HorDistRoad","Hillshade09","Hillshade12","Hillshade15",
           "HorDistFire")
# Four binary attributes for wilderness areas:
names <- c(names,"WA_RWA","WA_NWA","WA_CPWA","WA_CLPWA")
# 40 (!) binary attributes for soil types:
names <- c(names,sprintf("ST%02d",1:40))
# The cover type
names <- c(names,"Class")
# Assign these names to the attributes
names(covtype) <- names
# Let's also assign labels to the coverage types.
covtype$Class <- as.factor(covtype$Class)
levels(covtype$Class) <- c("Spruce/Fir", "Lodgepole Pine",
                           "Ponderosa Pine","Cottonwood/Willow","Aspen",
                           "Douglas-fir","Krummholz")
# How does it looks like?
str(covtype)


## ----covtype_wilder1-----------------------------------------------------
countWA <- covtype$WA_RWA + covtype$WA_NWA + covtype$WA_CPWA + covtype$WA_CLPWA
paste(min(countWA),max(countWA))


## ----covtype_soil1-------------------------------------------------------
soilT <- sprintf("ST%02d",1:40)
countST <- rowSums(covtype[,soilT])
paste(min(countST),max(countST))


## ----covtype_convert1----------------------------------------------------
# Which are the columns we want to consider?
searchOn <- c("WA_RWA","WA_NWA","WA_CPWA","WA_CLPWA")
# Get the index of each WA_
indexOfWA <- apply(covtype[,searchOn], 1, function(x) which(x == 1))
# Convert it to a factor with the column names we used.
factorOfWA <- factor(indexOfWA,labels = searchOn)
# Add it to the data frame
covtype$WildArea <- factorOfWA
# Drop the binary variables we don't need anymore
covtype[searchOn] <- list(NULL) 


## ----covtype_convert2----------------------------------------------------
# Which are the columns we want to consider?
searchOn <- sprintf("ST%02d",1:40)
searchOn
# Get the index of 1 in ST01..ST40
indexOfST <- apply(covtype[,searchOn], 1, function(x) which(x == 1))
# Convert it to a factor with the column names we used.
factorOfST <- factor(indexOfST,labels = searchOn)
# Add it to the data frame
covtype$SoilType <- factorOfST
# Drop the binary variables we don't need anymore
covtype[searchOn] <- list(NULL) 
str(covtype)


## ----covtype_summary-----------------------------------------------------
summary(covtype)
# Let's save it just in case!
save(covtype,file="Data/CoverageType/CovType2.R")


## ----covtype_boxplot1, fig.width=7, fig.height=7-------------------------
par(mfrow=c(1,3),mar=c(6,3,2,1))
boxplot(covtype$Elevation, main="Elevation",las=2)
boxplot(covtype$Slope, main="Slope",las=2)
boxplot(covtype$HorDistFire, main="Hor.Dist.Fire",las=2)


## ----covtype_boxplot2, fig.width=7, fig.height=7-------------------------
par(mfrow=c(1,3),mar=c(6,3,2,1))
boxplot(covtype$HorDistToHydro, main="Hor.Dist.Hydro",las=2)
boxplot(covtype$VertDistToHydro, main="Vert.Dist.Hydro",las=2)
boxplot(covtype$HorDistRoad, main="Hor.Dist.Road",las=2)


## ----covtype_viol1, fig.width=7, fig.height=7----------------------------
library(beanplot)
beanplot(covtype$HorDistToHydro,covtype$HorDistRoad,covtype$HorDistFire,
        names=c("Hor.Dist.Hydro","Hor.Dist.Road","Hor.Dist.Fire"))


## ----covtype_edatable1, results='asis'-----------------------------------
library(xtable)
tabSW <- table(covtype$SoilType,covtype$WildArea)
print(xtable(tabSW), type="html")


## ----covtype_edatable2, results='asis'-----------------------------------
tabCW <- table(covtype$Class,covtype$WildArea)
print(xtable(tabCW), type="html")


## ----covtype_edatable1v, fig.width=7, fig.height=8-----------------------
colfunc <- colorRampPalette(c("red", "yellow","green","cyan","blue","magenta"))
mosaicplot(t(tabSW),main="Soil Type x Wilderness Area",
              ylab="Soil Type",xlab="Wilderness Area",col=colfunc(nlevels(covtype$SoilType)))


## ----covtype_edatable2v, fig.width=7, fig.height=7-----------------------
mosaicplot(t(tabCW),main="Class x Wilderness Area",
              ylab="Class",xlab="Wilderness Area",col=colfunc(nlevels(covtype$Class)))