[R] Labels on a pyramide

[David Winsemius]

I took a shot at getting "assymetric lableling but it's by no means  
perfect. You really, really ,really, should learn to offer data with  
dput. See below:


On Feb 13, 2010, at 9:27 AM, Orvalho Augusto wrote:

> I am using pyramid.plot() from the plotrix package.
>
> I have something like this
>
> ############################################
> xy.pop<-dados$masfr
> xx.pop<-dados$femfr
> #agelabels<-dados$femlab
> xycol<-color.gradient(c(0,0,0.5,1),c(0,0,0.5,1),c(1,1,0.5,1),11)
> xxcol<-color.gradient(c(1,1,0.5,1),c(0.5,0.5,0.5,1),c(0.5,0.5,0.5,1), 
> 11)
> xylab<-dados$maslab
> xxlab<-dados$femlab
> agelabels<-xylab
>
> png("piramide9808.png")
> par 
> (mar 
> = 
> pyramid 
> .plot 
> (xy 
> .pop,xx.pop,labels=agelabels,top.labels=c("Masculino","","Feminino"),
>       main="Primeiras 10 cancros mais frequentes por
> sexo...",xycol=xycol,xxcol=xxcol,gap=0, labelcex=0))
>
> dev.off()
>
> #################################
>
> the dados a dataframe fabircated by someother program and looks like:
>
> ######################
> ordem	femlab	femfa	femfr	maslab	masfa	masfr
>
> 1	Colo do utero	258	26.76348548	Prostata	613	43.81701215
>
> 2	Mama	110	11.41078838	Figado	84	6.004288778
>
> 3	Esofago	62	6.43153527	Pele	70	5.003573981
>
> 4	Figado	60	6.22406639	Sarcoma de Kaposi	65	4.64617584
>
> 5	Pele	48	4.979253112	Esofago	63	4.503216583
>
> 6	Bexiga	37	3.838174274	Pulmao	46	3.288062902
>
> 7	Corpo do utero	34	3.526970954	Bexiga	43	3.073624017
>
> 8	"Utero, SOE"	28	2.904564315	Penis	33	2.358827734
>
> 9	Sarcoma de Kaposi	28	2.904564315	Laringe	27	1.929949964
>
> 10	Vulva e Vagina	24	2.489626556	Colon	24	1.715511079
>
> 11	Outras localizacoes	275	28.52697095	Outras localizacoes	331	 
> 23.65975697


In case anyone wants to take a crack at this without the hassle of  
recreating htis dataset...

dput(dados) can be used to recreate the complete dataframe

dados <- structure(list(ordem = 1:11, femlab = structure(c(2L, 6L, 4L,
5L, 8L, 1L, 3L, 10L, 9L, 11L, 7L), .Label = c("Bexiga", "Colo do utero",
"Corpo do utero", "Esofago", "Figado", "Mama", "Outras localizacoes",
"Pele", "Sarcoma de Kaposi", "Utero SOE", "Vulva e Vagina"), class =  
"factor"),
     femfa = c(258L, 110L, 62L, 60L, 48L, 37L, 34L, 28L, 28L,
     24L, 275L), femfr = c(26.76348548, 11.41078838, 6.43153527,
     6.22406639, 4.979253112, 3.838174274, 3.526970954, 2.904564315,
     2.904564315, 2.489626556, 28.52697095), maslab = structure(c(9L,
     4L, 7L, 11L, 3L, 10L, 1L, 8L, 5L, 2L, 6L), .Label = c("Bexiga",
     "Colon", "Esofago", "Figado", "Laringe", "Outras localizacoes",
     "Pele", "Penis", "Prostata", "Pulmao", "Sarcoma de Kaposi"
     ), class = "factor"), masfa = c(613L, 84L, 70L, 65L, 63L,
     46L, 43L, 33L, 27L, 24L, 331L), masfr = c(43.81701215, 6.004288778,
     5.003573981, 4.64617584, 4.503216583, 3.288062902, 3.073624017,
     2.358827734, 1.929949964, 1.715511079, 23.65975697)), .Names =  
c("ordem",
"femlab", "femfa", "femfr", "maslab", "masfa", "masfr"), class =  
"data.frame", row.names = c(NA,
-11L))



xy.pop<-dados$masfr
xx.pop<-dados$femfr
#agelabels<-paste(dados$femlab, "\t\t\t", dados$maslab, sep='')
xycol<-color.gradient(c(0,0,0.5,1),c(0,0,0.5,1),c(1,1,0.5,1),11)
xxcol<-color.gradient(c(1,1,0.5,1),c(0.5,0.5,0.5,1),c(0.5,0.5,0.5,1),11)
#xylab<-dados$maslab
#xxlab<-dados$femlab
#agelabels<-xylab


par(mar=pyramid.plot(xy.pop, xx.pop, labels= rep("",length(xy.pop)),  
top.labels=c("Masculino", "", "Feminino"),
       main="Primeiras 10 cancros mais frequentes por
sexo...", xycol=xycol,xxcol=xxcol, gap=0))
text(-xy.pop-5*nchar(dados$maslab), 1:length(xy.pop), dados 
$maslab );text(xx.pop+7*nchar(dados$femlab), 1:length(xx.pop), dados 
$femlab )

# Could not get the right ratio of xx.pop to nchar to get completely  
correct placement and given the  small number you might just want to  
put in a "hand-crafted" vector,

>
>
> #####################################
>
> The problem is (1) I do not want plot agelabels on the center and (2)
> I want plot different labels for each pair of the bars (one label for
> masculine and the other feminine).
>
> The data represent the 10 most frequent cancer in a group of  
> individuals.
>
> Can some one help please?
>
> Caveman
>
>
>
> -- 
> OpenSource Software Consultant
> CENFOSS (www.cenfoss.co.mz)
> SP Tech (www.sptech.co.mz)
> email: orvaquim at cenfoss.co.mz
> cell: +258828810980
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

David Winsemius, MD
Heritage Laboratories
West Hartford, CT