The folks over at the
molecular Ecologist have a instructive tutorial on map making using R. The data from the
birders in India, provides a novel dataset to try out these new map making skills. To start of with something simple, we look at the distribution of the 6 most sighted birds across India.
Only 6 species of birds have more than 700 recorded observations. The "Rosy Starling" leads the record with 864 observations followed by "Grey Wagtail" (857), "Common Sandpiper"(787), "Barn Swallow" (752),"Pied Cuckoo" (750) and "Greenish Warbler" with 710 observations.
read.csv("migrantwatch_reports.csv",row.names=NULL)->M
colnames(M)<-c("Species","Location name","City","State","Reporter","Date","Sighting type","Observation frequency","Start date","On behalf of","Latitude","Longitude")
numberofclusters<-10
N<-data.frame(M$Latitude,M$Longitude)
kmeans(N, numberofclusters)->P
M$cluster<-P$cluster
as.data.frame(P$centers)$M.Latitude->lat
as.data.frame(P$centers)$M.Longitude->lon
BSwallow<-as.data.frame(table(M$cluster[M$Species=="Barn Swallow"]))$Freq
CSandpiper<-as.data.frame(table(M$cluster[M$Species=="Common Sandpiper"]))$Freq
GWarbler<-as.data.frame(table(M$cluster[M$Species=="Greenish Warbler"]))$Freq
GWagtail<-as.data.frame(table(M$cluster[M$Species=="Grey Wagtail"]))$Freq
PCuckoo<-as.data.frame(table(M$cluster[M$Species=="Pied Cuckoo"]))$Freq
RStarling<-as.data.frame(table(M$cluster[M$Species=="Rosy Starling"]))$Freq
library(maps)
library(mapdata)
library(mapplots)
jpeg("top6.map.jpeg")
map("worldHires","India",col="gray90", fill=TRUE)
for (i in 1:numberofclusters) {
add.pie(z=c(BSwallow[i],CSandpiper[i],GWarbler[i],GWagtail[i],PCuckoo[i],RStarling[i]), x=lon[i], y=lat[i], col=c("blue","brown","Green","Grey","black","yellow"), labels="")
}
legend("topright",c("Barn Swallow","Common Sandpiper","Greenish Warbler","Grey Wagtail","Pied Cuckoo","Rosy Starling"),col=c("blue","brown","Green","Grey","black","yellow"),pch="*")
Running the above code gives a nice(take up the politics with the guys who made the R package) map with pie-charts that looks below image:
The pie-charts are generated for the mean locations that are identified by k-means clustering on all observations with a K of 10. So, this also gives us an idea of where most of the observations come from.
Rosy Starling's are definitely observed more often on the west coast while the Grey Wagtail is seen more often to the south. A large number (337) of the Rosy Starling observations come from
Shantilal Varu, whose favorite birds are waders.
