Data at the level of Gouvernorats

• We need first to use fortify from ggplot2 package and we will merge the new fortified data with the data containing the variable of interest and the HASC_1 variable.

> library(ggplot2)
> tn_gouv_fr <- fortify(m_gouv2,region = "HASC_1")

> i=match(tn_gouv_fr$id,dt1$HASC_1)
> tn_gouv_fr$IDR=dt1$IDR[i]
> colnames(tn_gouv_fr)
[1] "long"  "lat"   "order" "hole"  "piece" "id"    "group" "IDR"  

• Here’s then choropleth map:

>  p2<-ggplot(tn_gouv_fr, aes(x=long, y=lat, group=group)) +
+     geom_polygon(aes(fill=IDR),color = "black")+
+     labs(x="",y="")+ theme_bw()+
+     coord_fixed()
> p2<-p2+theme(axis.line=element_blank(),
+      axis.text.x=element_blank(),
+        axis.text.y=element_blank(),
+        axis.ticks=element_blank(),
+        axis.title.x=element_blank(),
+        axis.title.y=element_blank(),
+      panel.grid.major = element_blank(),
+      panel.grid.minor = element_blank(),
+      panel.border = element_blank(),
+      panel.background = element_blank())
>  p2<-p2+ theme(legend.position="right")
>  p2

Data at the level of delegations

• I will consider a data available in my Kaggle account where you can find several statistics about Tunisia at the level of delegations. You can find this data in the following link: https://www.kaggle.com/dhafer/eco-and-politics-data-on-tunisian-delegations. We will be interested here to display the difference of votes for Nahdha party between the Elections 2014 and 2011. We will consider a new indicator defined by \[\Delta\mbox{votes}=\displaystyle\frac{\mbox{Votes}(2011)-\mbox{Votes}(2014)}{\mbox{Votes}(2011)}\times 100\] It indicates the percentage of votes lost or gain from 2011 to 2014.

> library(readr)
> dt2<-read_csv("data_delegation_tunisia.csv")

> dt2=as.data.frame(dt2)
> head(dt2[,1:3])
  Gouvernorat      delegation   HASC_2
1      Ariana    Ariana-Ville TN.AN.AR
2      Ariana      Ettadhamen TN.AN.ET
3      Ariana Kalaat-Landlous TN.AN.KA
4      Ariana       La-Soukra TN.AN.LS
5      Ariana          Mnihla TN.AN.MN
6      Ariana          Raoued TN.AN.RA

• We load then the map at the delegation level and this time we should look to the HASC_2 variable

> library(raster)
> m_deleg2<- getData(name="GADM",  country="TUN", level=2)

> m_deleg2@data$HASC_2[1:4]
[1] "TN.AN.AR" "TN.AN.ET" "TN.AN.KA" "TN.AN.MN"

• We compute the new variable

> x=(dt2$Nahdha_2011-dt2$Nahdha_2014)/dt2$Nahdha_2011*100
> summary(x)
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
 -36.59   26.46   38.76   33.47   45.21   65.29       3 

• We use then fortify function and bind the new column to the data that will be displayed

> library(ggplot2)
> tn_deleg_fr <- fortify(m_deleg2,region = "HASC_2")

> i=match(tn_deleg_fr$id,dt2$HASC_2)
> tn_deleg_fr$DeltaVotes=x[i]

• Here’s the map then and we will add title to the legend and change the color when we fill the map. Negative differences (gain of votes) will be positive and Positive differences (lost of votes) will be represented in red.

> library(scales)
>  p2<-ggplot(tn_deleg_fr, aes(x=long, y=lat, group=group)) +
+     geom_polygon(aes(fill=DeltaVotes),color = "black")+
+     labs(x="",y="")+ theme_bw()+
+     coord_fixed()
> p2<-p2+scale_fill_gradientn(colours=c("blue","white","red"), 
+     values=rescale(c(-100,0,100)))
> 
> p2<-p2+theme(axis.line=element_blank(),
+      axis.text.x=element_blank(),
+        axis.text.y=element_blank(),
+        axis.ticks=element_blank(),
+        axis.title.x=element_blank(),
+        axis.title.y=element_blank(),
+      panel.grid.major = element_blank(),
+      panel.grid.minor = element_blank(),
+      panel.border = element_blank(),
+      panel.background = element_blank())
>  p2<-p2+ theme(legend.position="right")
>  p2+labs(fill = "% of the Diff, \n  of Votes \n 2011 - 2014")