Mª Teresa Martinez Soriano
2013-Nov-11 22:52 UTC
[R] colours legend, for loop,density plot
Hi , thanks in advance I have the follow code: normal<-sort(rnorm(1000)) cauchy<-sort(rcauchy(1000)) t3<-sort(rt(1000,3)) t10<-sort(rt(1000, 10)) col<-c("green","blue","orange","purple") v<-list(normal,cauchy,t3,t10) names(v)<-c("Normal", "Cauchy", "T-stud 3 df", "T-stud 10 df") par(mfrow=c(1,2)) plot(density(normal),col=col[[1]],main="Funciones de densidad") for ( i in 2:4) { lines(density(v[[i]]),col=col[[i]],lty=i+2) } legend(x=-4,y=0.3,names(v),col=col,cex=0.6) The problem is that in the legend doesn't appear colours so I can not identify which curve is each one, please could you tell me what do I neet to change in order to solve it?? Thanks a lot, Tere [[alternative HTML version deleted]]
On 11/12/2013 09:52 AM, M? Teresa Martinez Soriano wrote:> normal<-sort(rnorm(1000)) cauchy<-sort(rcauchy(1000)) t3<-sort(rt(1000,3)) t10<-sort(rt(1000, 10)) > col<-c("green","blue","orange","purple") v<-list(normal,cauchy,t3,t10) names(v)<-c("Normal", "Cauchy", "T-stud 3 df", "T-stud 10 df") > par(mfrow=c(1,2)) plot(density(normal),col=col[[1]],main="Funciones de densidad") for ( i in 2:4) { lines(density(v[[i]]),col=col[[i]],lty=i+2) } > legend(x=-4,y=0.3,names(v),col=col,cex=0.6) >Hi Tere, Try this: legend(x=-2,y=0.04,names(v),col=col,cex=0.6,lty=c(1,4:6)) Jim
Mª Teresa Martinez Soriano
2013-Dec-02 11:03 UTC
[R] Confidence interval, multiple imputation
Hi to everyone, I have a big data set where rows are observations and columns are variables. It contains a lot of missing values. I have used multiple imputation with library mice and I get an “exact” prediction of each missing value. Now, I would like to know the error I can commit or the confidence interval. How can I get this? This is part of my code library(mice) mod1<-mice(dat, method=c("","",rep("pmm",6))) ro<-round(cor(dat, use = "pair"), 3) predictor<-quickpred(dat)# esta matriz predictora se construye según las correlaciones mod1<-mice(dat,method=c("","",rep("pmm",6)), pred=predictor) imputados<-complete(mod1,'long') x.imp=split(imputados, imputados$.imp) acumula=x.imp[[1]][,-c(1,2)] for(j in 2:length(x.imp)) { acumula=acumula+x.imp[[j]][,-c(1,2)]} med.imp=acumula/5 Thanks in advance [[alternative HTML version deleted]]