R help - May 2012 - Correlograms: using boxes and different variables on rows and columns

I'm trying to make correlograms using corrgram.  See below for a simple
example.

####
library(corrgram)

data(baseball)
vars1 <-
c("Assists","Atbat","Errors","Hits","Homer","logSal")
vars2 <-
c("Putouts","RBI","Runs","Walks","Years")

corrgram(baseball[,vars2],lower.panel=panel.shade, upper.panel=panel.pie)
####

I am having two problems:

1) I want to change the visual to "bars" (see Figure 1 on p. 3:
http://www.datavis.ca/papers/corrgram.pdf).  I tried changing the code for
the upper panel to include bars, but it doesn't appear correctly:

corrgram(baseball[,vars2],lower.panel=panel.shade, upper.panel=panel.bars)

2) I want to display one set of variables in the columns (vars1) and another
set of variables in the rows (vars2).  The reason I want to keep one set of
variables in the columns and another set of variables in the rows is because
I have many variables and am only interested in the correlations of vars1
with vars2 (and not of the intercorrelations of variables among vars1 or
among vars2).  Thus, the correlogram with separate variables in the rows and
columns would be more condensed and easier to examine visually.

Can I address these issues in corrgram, or is there another package that
will allow me to do that?  Thanks in advance.

--
View this message in context:
http://r.789695.n4.nabble.com/Correlograms-using-boxes-and-different-variables-on-rows-and-columns-tp4631309.html
Sent from the R help mailing list archive at Nabble.com.

In general, when you have a question about a package, it is best to
contact the package author directly.  (In this case, me).

1. Easy.  You just have to define your own panel function.  I just
modified panel.shade to create panel.bar

panel.bar <- function(x, y, corr=NULL, ...){

  usr <- par()$usr
  minx <- usr[1]; maxx <- usr[2]
  miny <- usr[3];  maxy <- usr[4]

  if (is.null(corr))
    corr <- cor(x, y, use = "pair")
  ncol <- 14
  pal <- col.corrgram(ncol)
  col.ind <- as.numeric(cut(corr, breaks = seq(from = -1, to = 1,
                                    length = ncol + 1), include.lowest = TRUE))
  col.bar <- pal[col.ind]
  if(corr < 0) {
    maxy <- miny + (maxy-miny) *  abs(corr)
    rect(minx, miny, maxx, maxy, col = pal[col.ind],
         border = "lightgray")
  } else if (corr > 0){
    miny <- maxy - (maxy-miny)*corr
    rect(minx, miny, maxx, maxy, col = pal[col.ind],
         border = "lightgray")
  }

}

corrgram(auto, order=TRUE, main="Auto data (PC order)",
upper.panel=panel.bar)

2.  Harder.  You need to create your own copy of the corrgram function
and change it.  Look for the call to par() and change the number of
rows and columns.   Then change the loops over i and j to include only
the row and column numbers that you want instead of all rows and
columns.

Kevin

On Fri, May 25, 2012 at 8:38 AM, dadrivr <dadrivr at gmail.com>
wrote:
> I'm trying to make correlograms using corrgram. ?See below for a simple
> example.
>
> ####
> library(corrgram)
>
> data(baseball)
> vars1 <-
c("Assists","Atbat","Errors","Hits","Homer","logSal")
> vars2 <-
c("Putouts","RBI","Runs","Walks","Years")
>
> corrgram(baseball[,vars2],lower.panel=panel.shade, upper.panel=panel.pie)
> ####
>
> I am having two problems:
>
> 1) I want to change the visual to "bars" (see Figure 1 on p. 3:
> http://www.datavis.ca/papers/corrgram.pdf). ?I tried changing the code for
> the upper panel to include bars, but it doesn't appear correctly:
>
> corrgram(baseball[,vars2],lower.panel=panel.shade, upper.panel=panel.bars)
>
> 2) I want to display one set of variables in the columns (vars1) and
another
> set of variables in the rows (vars2). ?The reason I want to keep one set of
> variables in the columns and another set of variables in the rows is
because
> I have many variables and am only interested in the correlations of vars1
> with vars2 (and not of the intercorrelations of variables among vars1 or
> among vars2). ?Thus, the correlogram with separate variables in the rows
and
> columns would be more condensed and easier to examine visually.
>
> Can I address these issues in corrgram, or is there another package that
> will allow me to do that? ?Thanks in advance.
>
> --
> View this message in context:
http://r.789695.n4.nabble.com/Correlograms-using-boxes-and-different-variables-on-rows-and-columns-tp4631309.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> 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.


-- 
Kevin Wright