location.only.R

library("dplyr")
library("maps")
library("ggplot2")
library("mapproj")

theme_clean <- function(base_size = 12) {
  require(grid)
  theme_grey(base_size) %+replace%
    theme(
      axis.title      =   element_blank(),
      axis.text       =   element_blank(),
      panel.background    =   element_blank(),
      panel.grid      =   element_blank(),
      axis.ticks.length   =   unit(0,"cm"),
      axis.ticks.margin   =   unit(0,"cm"),
      panel.margin    =   unit(0,"lines"),
      plot.margin     =   unit(c(0,0,0,0),"lines"),
      legend.position = "bottom",
      complete = TRUE
    )
}

simpleCap <- function(x) {
  s <- strsplit(x, " ")[[1]]
  paste(toupper(substring(s, 1,1)), tolower(substring(s, 2)),
        sep="", collapse=" ")
}
vsimpleCap=Vectorize(simpleCap)

state_map=map_data(map="state")
df=read.csv("directory.csv.gz", header=TRUE,stringsAsFactors=FALSE)
st=read.csv("state_table.csv.gz", header=TRUE,stringsAsFactors=FALSE)
names(st)=c("name","state","USregion")
df=left_join(df,st,by="state")
rg=read.csv("regions.csv.gz", header=TRUE,stringsAsFactors=FALSE)

#Cleanup
df=mutate(df,
          country=vsimpleCap(country),
          country=ifelse(country=="","USA",as.character(country)),
          country=ifelse(country=="United Kingdom","UK",as.character(country)),
          country=ifelse(country=="Uk","UK",as.character(country)),
          country=ifelse(country=="P.r. China","China",as.character(country)),
          country=ifelse(country=="Columbia","Colombia",as.character(country)),
          country=ifelse(country=="Republic Of Armenia","Armenia",as.character(country))
)

countries=df %>% group_by(country) %>% summarise(count=length(country)) %>% arrange(-count) 
countries=left_join(countries,rg)
countries$region=factor(countries$region,levels=(countries%>%group_by(region)%>%summarize(count=sum(count))%>%arrange(-count))$region)
countries$usaqual = factor(with(countries,ifelse(country=="USA","USA","Rest of World")),levels=c("USA","Rest of World"))
usaness= countries %>% group_by(usaqual) %>% summarize(count=sum(count)) %>% arrange(-count)

p=ggplot(usaness,aes(x=usaqual,y=count,fill=usaqual))
p=p+geom_bar(stat="identity") 
p=p+labs(x="",y="",title="Members in USA and Rest of World")
p=p+theme(legend.position = "none")
p
ggsave(plot=p,file="USAvRest.png",width=4,height=4)

p=ggplot(countries,aes(x=region,y=count,fill=region))
p=p+geom_bar(stat="identity") 
p=p+labs(x="",y="",title="Members by Region")
p=p+theme( axis.text.x = element_text(angle=90, vjust=0.5),legend.position="none")
p=p+theme(legend.position = "none")
p
ggsave(plot=p,file="MembersByRegion.png",width=4,height=4)

p=ggplot(filter(countries,country!="USA" & count>5),aes(x=count,y=reorder(country,count)))
p=p+geom_point(size=3) 
p=p+labs(x="Members",y="",title="Members Outside USA\nin Countries with >5 Members")
p=p+theme_bw()
p
ggsave(plot=p,file="MembersByCountry.png",width=4,height=6)

#Make state map
states=df %>% filter(!is.na(USregion)) %>% group_by(name) %>% summarise(count=length(state)) %>% arrange(-count) 
states$name=tolower(states$name)
states[states$name=="washington dc","name"]="district of columbia"
allst=data.frame(name=as.character(unique(state_map$region)))
states=merge(allst,states,by="name",all.x=TRUE) %>% arrange(-count) 
states[is.na(states$count),2]=0
names(states)=c("state","count")

p=ggplot(states, aes(map_id=state, fill=count))
p=p+geom_map(map=state_map,color="black")
p=p+scale_fill_gradient2(low="#559999",mid="grey90",high="#BB650B",midpoint=median(states$count))
p=p+expand_limits(x=state_map$long,y=state_map$lat)
p=p+coord_map("polyconic")
p=p+theme_clean()
p
ggsave(plot=p,file="us.member.map.png",width=6,height=4)

states$state_simplecap = vsimpleCap(as.character(states$state))
#Members in states with > 5 members
p=ggplot(filter(states, count>5),aes(x=count,y=reorder(state_simplecap,count)))
p=p+geom_point(size=3) 
p=p+labs(x="Members",y="",title="Members in States \nWith >5 Members")
p=p+theme_bw()
p
ggsave(plot=p,file="MembersByState.png",width=4,height=8)