How to remove spacing in pie charts on multiple panels and have only one legend on top using r

Question

Pie Charts:

DATA LINK: https://drive.google.com/file/d/0BwoPt0jyGdzORkM3cVA0WjJodVk/view?usp=sharing

Mydata<-read.csv(file="final_analysis_candy_analysis.csv",head=TRUE,sep=",") dhfr.Arg <- table(Mydata$dhfr.Arg.59.163.137.) dhfr.Ile <- table(Mydata$dhfr.Ile.51.214.65.) dhfr.Asn108 <- table(Mydata$dhfr.Asn108.328.372.) glu.540 <- table(Mydata$glu.540.538.326.200.) gly.437 <- table(Mydata$gly.437.848.300.) library(plotrix) par(op) op <-par(mfrow=c(2,3),mar=c(0,0,1,0)) pct <- round(dhfr.Arg/sum(dhfr.Arg)*100) lbls <- paste(names(dhfr.Arg), pct) # add percents to labels lbls <- paste(lbls,"%",sep="") # ad % to labels lp<-pie3D(dhfr.Arg,radius=0.8,labels=lbls,explode=0.1, labelrad=1.4,main="dhfr Arg 59(163,137)") pct <- round(dhfr.Ile/sum(dhfr.Ile)*100) lbls <- paste(names(dhfr.Ile), pct) # add percents to labels lbls <- paste(lbls,"%",sep="") # ad % to labels lp<-pie3D(dhfr.Ile,radius=0.8,labels=lbls,explode=0.1, labelrad=1.4,main="dhfr Ile 51(214,65)") pct <- round(dhfr.Asn108/sum(dhfr.Asn108)*100) lbls <- paste(names(dhfr.Asn108), pct) # add percents to labels lbls <- paste(lbls,"%",sep="") # ad % to labels lp <- pie3D(dhfr.Asn108,radius=0.8,labels=lbls,explode=0.1, labelrad=1.4,main="dhfr Asn108(328,372)") pct <- round(glu.540/sum(glu.540)*100) lbls <- paste(names(glu.540), pct) # add percents to labels lbls <- paste(lbls,"%",sep="") # ad % to labels lp<-pie3D(glu.540,radius=0.8,labels=lbls,explode=0.1, labelrad=1.4,main="glu 540(538,326,200)") pct <- round(gly.437/sum(gly.437)*100) lbls <- paste(names(gly.437), pct) # add percents to labels lbls <- paste(lbls,"%",sep="") # ad % to labels lp <- pie3D(gly.437,radius=0.8,labels=lbls,explode=0.1, labelrad=1.4,main="gly 437(848,300)") par(op) 

Answer 1

Using 2D visualization, your plot is much easier to understand. Therefore an alternative two-dimensional solution with ggplot :

 # load needed packages library(data.table) library(ggplot2) library(scales) # process & summarise the data (with data.table) mydat <- melt(setDT(Mydata), id=1, measure.vars=4:8)[, .N, by = .(variable,value) ][, `:=` (perc = round(N/sum(N),2), pos = cumsum(N)-0.5*N), by = variable] # create the plot with ggplot2 & scales ggplot(mydat) + geom_bar(stat="identity", aes(x="", y=N, fill=value)) + geom_text(aes(x = "", y = pos, label = percent(perc))) + scale_x_discrete(expand = c(0,0)) + scale_y_continuous(expand = c(0,0)) + coord_polar(theta = "y") + facet_grid(.~ variable, scales = "free") + theme_minimal(base_size = 16) + theme(axis.title = element_blank(), axis.text = element_blank(), panel.grid = element_blank(), legend.title = element_blank()) 

which gives the following graph:

See this answer on how to calculate pos variable with base R or plyr and dplyr packages.

---

However, pie charts are generally not the best way to visualize data. Also in this case, the bart chart will result in clearer visualization. WITH:

 ggplot(mydat, aes(x=variable, y=perc, fill=value)) + geom_bar(stat="identity", aes(label = percent(perc)), width=0.6) + scale_y_continuous(labels = percent(c(0,0.25,0.50,0.75,1.00))) + coord_flip() + theme_minimal(base_size = 14) + theme(axis.title = element_blank(), legend.title = element_blank()) 

You are getting:

Answer 2

From plotrix documentation there is a margin mar parameter - fields around the pie. See the default value is mar=c(4,4,4,4) , which creates this space.

 pie3D(x,edges=NA,radius=1,height=0.1,theta=pi/6,start=0,border=par("fg"), col=NULL,labels=NULL,labelpos=NULL,labelcol=par("fg"),labelcex=1.5, labelrad=1.25,sector.order=NULL,explode=0,shade=0.8,mar=c(4,4,4,4),pty="s",...) 

Try setting mar to lower amounts, so add this parameter to your pie3D, for example. mar =c(1,1,1,1) to omit all fields.