[R] Multiple R linear models into one Latex table

[Matevž Pavlič]

Sorry wrong thread ;)

m

-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On Behalf Of Matevž Pavlič
Sent: Monday, August 22, 2011 12:14 PM
To: Rmh; Alex Ruiz Euler
Cc: r-help at r-project.org
Subject: Re: [R] Multiple R linear models into one Latex table

Hi, thanks for the help. 

Class says as follows :

> class(DF)
[1] "SpatialPointsDataFrame"
attr(,"package")
[1] "sp"
> class(grd)
[1] "SpatialPixels"
attr(,"package")
[1] "sp

Anyway, the problem was, as Rmh suggested, with the zerodist().

Tnx, m

-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On Behalf Of Rmh
Sent: Monday, August 22, 2011 11:15 AM
To: Alex Ruiz Euler
Cc: r-help at r-project.org
Subject: Re: [R] Multiple R linear models into one Latex table

please look at the latex() function in package "Hmisc".

Sent from my iPhone

On Aug 22, 2011, at 0:55, Alex Ruiz Euler <rruizeuler at ucsd.edu> wrote:

> 
> 
> Dear community,
> 
> I had been looking for an easy way to produce latex tables from R 
> output. xtable() and the package apsrtable produce good outputs but 
> they are not exactly what I was looking for.
> 
> I wrote this code that generates regression tables from multiple R 
> linear models. I want to share it because it might be useful for 
> someone else, and because I would appreciate comments on how to 
> optimize the code. I'm also open to suggestions about design or packages to include. Or pointing out bugs. I hope this is the right list to post.
> 
> Please consider it is still in a very primitive form, and the generated table might look strange for some disciplines (I come from political science).
> 
> Thanks,
> Alex
> 
> 
> 
> ## Start script
> 
> 
> #################################
> #### R models to Latex table ####
> #################################
> 
> 
> # nice() - Multiple R linear models into one Latex table.
> #
> #    nice(list(fit1,fit2,...,fitn)) for generic table with n models
> #
> #                    or
> #
> #  for better tables, create objects 'model.names' and 
> 'final.varnames', then #
> #    nice(list(fit1,fit2,...,fitn), model.names, final.varnames)
> 
> 
> 
> nice <- function(modelos, model.names=NULL, final.varnames=NULL) {
> 
>  var.names<-vector(mode="character")
>  k<-1
>  for (i in 1:length(modelos)) {
>    l<-length(names(coef(modelos[[i]])))
>    var.names[c(k:c(k+l-1))]<-names(coef(modelos[[i]]))
>    k<-k+l
>  }
>  var.names<-unique(var.names)
> 
>  if(is.null(final.varnames)) {
>  final.varnames<-var.names
>  }
> 
>  if(is.null(model.names)) {
>  model.names<-paste("Model",1:length(modelos))
>  }
> 
>  mat.all<-matrix(data=NA, nrow=length(var.names)*2, byrow=FALSE, 
> ncol=length(model.names), dimnames=list(c(rep(c("coef","sd"),length=length(var.names)*2)), c(model.names)))  dimnames(mat.all)[[1]][c(seq(1,dim(mat.all)[1], by=2))]<-var.names; dimnames(mat.all)[[1]][c(seq(2,dim(mat.all)[1], by=2))]<-""
> 
>  for (j in 1:length(modelos)) {
>    
> mat.all[c(match(names(coef(modelos[[j]])),dimnames(mat.all)[[1]])),j]<
> -coef(modelos[[j]])
>    
> mat.all[c(match(names(coef(modelos[[j]])),dimnames(mat.all)[[1]])+1),j
> ]<-sqrt(diag(vcov(modelos[[j]])))
>  }
> 
>  mat.all<-signif(mat.all, digits=3);
> dimnames(mat.all)[[1]][c(seq(1,dim(mat.all)[1],
> by=2))]<-final.varnames
> 
> 
>  cat("\\begin{table}[!hbt]", "\n","\\caption[Short for List of 
> Tables]{Long for this 
> title}","\n","\\begin{center}","\n","\\begin{tabular}{llllllllll}","\n
> ")
>  
> cat("&",paste(model.names[1:length(model.names)-1],"&"),model.names[le
> ngth(model.names)],"\\\\","\n","\\hline","\n")
>  for(i in 1:dim(mat.all)[1]) {
>    if(i %in% c(seq(1,dim(mat.all)[1], by=2))) { # Odd rows
> (coefficients)
> 
>  cat(dimnames(mat.all)[[1]][i],"&")
>  cat(
>    for (j in 1:(length(modelos)-1)) {
>      if(!is.na(mat.all[i,j])) {
>    p.val<-abs(mat.all[i,j]/mat.all[i+1,j])
>    if(p.val>1.644) {
>      if(1.645<=p.val & p.val<1.96) cat("\\textbf{",round(mat.all[i,j], digits=5),"*}&", sep="") else {
>        if(1.96<p.val & p.val<2.576) cat("\\textbf{",round(mat.all[i,j], digits=5),"**}&", sep="") else {
>          cat("\\textbf{",round(mat.all[i,j], digits=5),"***}&", sep="")
>        }
>      }
>      } else cat(round(mat.all[i,j], digits=5),"&")
>    } else cat("&")
>  })
> 
>  cat(
>    for (j in length(modelos) ) {
>      if(!is.na(mat.all[i,j])) {
>    p.val<-abs(mat.all[i,j]/mat.all[i+1,j])
>    if(p.val>1.644) {
>      if(1.645<=p.val & p.val<1.96) cat("\\textbf{",round(mat.all[i,j], digits=5),"*}", sep="") else {
>        if(1.96<p.val & p.val<2.576) cat("\\textbf{",round(mat.all[i,j], digits=5),"**}", sep="") else {
>          cat("\\textbf{",round(mat.all[i,j], digits=5),"***}", sep="")
>        }
>      }
>    } else cat(round(mat.all[i,j], digits=5))
>    }
>   })
>  cat("\\\\", "\n" )
> 
>  } else {
>      cat("&",ifelse(!is.na(mat.all[i,1:(length(dimnames(mat.all)[[2]])-1)]),paste("(\\emph{",round(mat.all[i,1:(length(dimnames(mat.all)[[2]])-1)], digits=5),"})&",sep=""), paste("&")),
>       ifelse(!is.na(mat.all[i,length(dimnames(mat.all)[[2]])]),paste("(\\emph{",round(mat.all[i,length(dimnames(mat.all)[[2]])], digits=5),"})",sep=""),paste(" ")),
>    "\\\\", sep="", fill=TRUE)}
>    }
> 
>  cat("\\hline","\n","\\emph{n}","&")
>  cat(
>      for(j in 1:(length(modelos)-1)) {
>    cat(length(modelos[[j]]$residuals),"&",sep="")
>      },
>      for(j in length(modelos)) {
>    cat(length(modelos[[j]]$residuals), "\\\\", "\n", sep="")
>      }
>  )
> 
>  cat("\\emph{Adj-R$^2$}&")
>  cat(
>    for(j in 1:(length(modelos)-1)) {
>    if (class(modelos[[j]])[1]=="lm") {cat(round(summary(modelos[[j]])$adj.r.squared,digits=2),"&", sep="")} else {
>    cat("&") 
>    }
>      },
>      for(j in length(modelos)) {
>    if (class(modelos[[j]])[1]=="lm") {cat(round(summary(modelos[[j]])$adj.r.squared,digits=2),"\\\\","\n")} else {
>    cat("\\\\","\n") 
>    }
>      } 
>    )
> 
>  cat("\\emph{AIC}&")
>  cat(
>    for(j in 1:(length(modelos)-1)) {
>    if (class(modelos[[j]])[1]=="glm") {cat(round(summary(modelos[[j]])$aic, digits=0),"&", sep="")} else {
>    cat("&")
>    }
>      },
>      for(j in length(modelos)) {
>    if (class(modelos[[j]])[1]=="glm") {cat(round(summary(modelos[[j]])$aic, digits=0),"\\\\", sep="")} else {
>    cat("\\\\") 
>    }
>      }
>    )
>  cat("\n","\\hline")
>  cat("\n","\\multicolumn{3}{l}{\\footnotesize{Signif. codes:  0.01 
> `***' 0.05 `**' 0.1 `*'}}\\\\") 
> cat("\n","\\multicolumn{3}{l}{\\footnotesize{Standard errors in
> parentheses}}\\\\")  cat("\n","\\end{tabular}","\n","\\end{center}",
> "\n", "\\end{table}", "\n") # cat("% Suggestions: Alex R.E. 
> <rruizeuler at ucsd.edu>","\n") }
> 
> 
> ## End script
> 
> ______________________________________________
> 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.

______________________________________________
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.

______________________________________________
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.