view scripts_R/Metadata_subsetBLSM.R @ 105:edd82eb89b4b branch-tests tip

Merge

df = read.csv('../data/results/df_and_clusters.csv', header=T)

source("MetadataPlots.R")

## for plos use arial
#install.packages("extrafont")
library(extrafont)
font_import()
loadfonts()
Arial <- Type1Font(family="Arial", metrics=c("ArialMT.afm","arial-BoldMT.afm","Arial-ItalicMT.afm", "Arial-BoldItalicMT.afm"))
postscriptFonts(Arial=Arial)
par(family="Arial")

#pdf(file="data/country_distribution_BL.pdf")
pdf(file="../data/results/country_distribution.pdf", pointsize=12, width=6, height=4)
PlotCountryCounts(df)
dev.off()
postscript(file="../data/results/country_distribution.eps", pointsize=12, width=6, height=4)
PlotCountryCounts(df)
dev.off()

pdf(file="../data/results/year_distribution.pdf", width=6, height=4, pointsize=12)
PlotYearDistribution(df)
dev.off()
postscript("../data/results/year_distribution.eps", width=10, pointsize=12)
PlotYearDistribution(df)
dev.off()
#PlotBarChart(df, cat="Year", ordercat="REGION", mincount=10)

#pdf(file="data/language_distribution_BL.pdf")
levels(df$Language)[which(levels(df$Language)=="Southwestern Caribbean Creole English")]="SouthW Carib. Creole English"
df$Language[which(df$Language=="Southwestern Caribbean Creole English")] = "SouthW Carib. Creole English"
levels(df$Language)[which(levels(df$Language)=="Lesser Antillean Creole French")]="Lesser Antil. Creole French"
df$Language[which(df$Language=="Lesser Antillean Creole French")] = "Lesser Antil. Creole French"
df$REGION[which(df$Country=="French Guiana")] = "South America"
pdf(file="../data/results/language_distribution.pdf", pointsize=12)
PlotBarChart(df, cat="Language", ordercat="Region", mincount=10)
dev.off()
postscript("../data/results/language_distribution.eps", width=8, height=10, pointsize=12)
PlotBarChart(df, cat="Language", ordercat="Region", mincount=10)
dev.off()

df = read.csv('../data/results/df_and_clusters.csv', header=T)
#PlotBarChart(df, cat="Clusters", ordercat="CountryLang", mincount=1,legend=F)
df$REGION[which(df$Country=="French Guiana")] = "South America"
g = ggplot(df,aes(df$Clusters, fill=df$REGION))+geom_bar()
levels(df$REGION)[which(levels(df$REGION)=="South America")]="S. America"
levels(df$REGION)[which(levels(df$REGION)=="North America")]="N. America"

#library(rworldmap)
#wrld = getMap()
#regiondata<-wrld@data[,c("ADMIN","GEO3", "Stern")]
#df<-merge(df,regiondata,by.x="Country",by.y="ADMIN",all.x=T)

#cluster_labels_df = read.csv('data/clusters_top3_labels.csv')
cluster_labels_df = read.csv('../data/results/clusters_top3_countries.csv')
cluster_labels = paste(cluster_labels_df[,1],cluster_labels_df[,2],cluster_labels_df[,3],sep="")
#df$CountryLang = as.factor(paste(df$Country, df$Language, sep="-"))

countrycounts = table(df$Clusters,df$Country)
library(cluster)
library(ape)
library(gridExtra)
library(ggdendro)
library(dendextend)
hc = hclust(dist(countrycounts), method="average")
hc2=hc
#hc2$labels = as.character(1:length(cluster_labels))
hc2$labels = ""
#dhc <- as.dendrogram(hc2)
# library(dynamicTreeCut)
# clusters <- cutreeDynamic(hc2, minClusterSize = k_clust,method = "tree")
# clusters <- clusters[order.dendrogram(dhc)]
# clusters_numbers <- unique(clusters) - (0 %in% clusters)
# n_clusters <- length(clusters_numbers)
# library(colorspace)
# cols <- rainbow_hcl(n_clusters)
# dhc <- hc2 %>% as.dendrogram %>%
#   set("branches_k_color", k=k_clust) %>% branches_attr_by_clusters(clusters, values = cols)
k_clust = 5
dhc <- hc2 %>% as.dendrogram %>%
  set("branches_k_color", k=k_clust) %>% set("branches_lwd", 0.7) %>% 
  set("labels_cex", 0.6) %>% set("labels_colors", k=k_clust) %>%
  set("leaves_pch", 19) %>% set("leaves_cex", 0.5) 
#ddata <- dendro_data(dhc, type = "rectangle")
ddata <- as.ggdend(dhc)
p <- ggplot(ddata)+coord_flip()
#p <- ggplot(segment(ddata)) + 
#  geom_segment(aes(x = x, y = y, xend = xend, yend = yend, colour=ddata$segments$col)) + 
#  coord_flip() + theme_dendro() + theme(legend.position="none") +
#  geom_text(aes(x = x, y = y, label = label, angle = -90, hjust = 0.5, vjust=1.3, colour=ddata$labels$col), data= label(ddata))

#dend <- hc2 %>% as.dendrogram %>%
#  set("branches_k_color", k = 5) %>% set("branches_lwd", 0.7) %>%
#  set("labels_cex", 0.6) %>% set("labels_colors", k = 5) %>%
#  set("leaves_pch", 19) %>% set("leaves_cex", 0.5) 
#ggd1 <- as.ggdend(dend)
#pp <- ggplot(ggd1, horiz = TRUE)

library(stringr)
for (i in 1:length(cluster_labels)){
  cl = cluster_labels[i]
  cl = str_replace_all(cl, "[(']", "")
  cl = str_replace_all(cl, "[|]", "-")
  cl = str_replace_all(cl, ", ", " (")
  cl = str_replace_all(cl, "[)]", "), ")
  cl = str_replace_all(cl, "nan", "NA")
  #cl = paste(cl, "cluster",i)
  cluster_labels[i] = cl
}
#cluster_idx = paste("cluster",1:length(cluster_labels))
#df$Clusters = as.factor(df$Clusters)
df$Clusters = factor(x=df$Clusters,levels=hc$labels[hc$order])
cluster_labels = cluster_labels[hc$order]
#g = ggplot(df,aes(as.factor(df$Clusters), fill=df$CountryLang))+geom_bar()
#g = ggplot(df,aes(Clusters, fill=REGION))+geom_bar()+facet_grid(~REGION,space="free",scales="free")#,scales="free")
g = ggplot(df,aes(as.factor(df$Clusters), fill=df$Region))+geom_bar()
#g = ggplot(df,aes(as.factor(df$Clusters), fill=df$REGION))+geom_bar()
g = g+scale_x_discrete(labels=cluster_labels)
#g = g+scale_y_continuous(position="right")
#g = g+scale_fill_brewer(palette="Paired")#+scale_fill_grey()
g = g+scale_fill_brewer(palette="Paired")#+scale_fill_grey()
#g = g+labs(y="Counts", x="Top 3 country-language tags in each cluster")+coord_flip()+theme_bw()#+guides(fill="none")
g = g+labs(y="Counts", x="Clusters")+coord_flip()+theme_bw()#+guides(fill="none")
#g = g+labs(y="Counts", x="Clusters")+coord_flip()+theme_bw()#+guides(fill="none")
#g = g+guides(fill = guide_legend(title = "Region"))+theme(legend.position=c(.9,.8),legend.margin = unit(0, "cm"),legend.key.size = unit(0.3, "cm"),legend.title = element_text(size=10),legend.text = element_text(size=10))
#g = g+guides(fill = guide_legend(title = "Region"))+theme(legend.position="left",legend.margin = unit(0, "cm"),legend.key.size = unit(0.3, "cm"),legend.title = element_text(size=9),legend.text = element_text(size=9))
g = g+guides(fill = guide_legend(title = "Region"))+theme(legend.position="top",legend.title = element_text(size=9),legend.text = element_text(size=9))
g = g+theme(panel.border = element_rect(colour = "white"),strip.background=element_rect(fill="white"),strip.text.x = element_blank())
#g = g+theme(axis.text.y = element_text(colour = ddata$labels$col))
ggsave('../data/results/clusters_top3.pdf',plot=g, pointsize=12)
ggsave('../data/results/clusters_top3.eps',plot=g, pointsize=12)