[R] some newbie problems with plotting and RPgSQL

Paul Murrell paul at stat.auckland.ac.nz
Wed Jul 18 01:23:26 CEST 2001


> I am using RPgSQL to grab some data from a postgresql database (about 179
> columns and a few thousand rows), and I want to do a lot of pairwise
> comparison of the rows...(this is yeast genomic microarray data for those
> who care)
> I get to the database:
> > db.connect(dbname="yeast")
> and grab the first two rows like this:
> > t<-sql.select(columns="*", from="gasch", limit=2)
> then create two lists a la:
> > a<-t[1,][4:length(t[1,])]
> > b<-t[2,][4:length(t[2,])]
> (the first 3 entries are text)
> and now I just want to see a plot(a,b) but I always get the following
> error:
> Error in plot.new() : Figure margins too large
> (I even tried par(mar=c(0,0,0,0)))
> Oddly I get the following behaviour:
> > plot(seq(1,174),a)
> works (the graph looks like it should)
> but the reverse
> > plot(a,seq(1,174))
> gives me:
> Error in plot.new() : Figure margins too large
> similarly for b...so I am guessing the problem is passing my lists as the
> first argument...
> Manually getting the rows and making a text file that explictly states the
> vectors and then sourcing the file  from R does plot(a,b) just fine....
> If I understand the R classes correctly t is a dataframe and a and b are
> lists (named by the column headings of the database)....

I don't think so.  A subset of the rows of a data.frame is still a
data.frame ...

    > temp <- data.frame(x=1:10, y=10:1)
    > class(temp[1,])
    [1] "data.frame"

This means that plot(a, <whatever>) or plot(b, <whatever>) is going to call
the plot.data.frame() method.  This is going to try to do a pairs() plot
with 174*174 plots :)
Hence the error message that there is not enough room to draw the plots.

Note that plot(1:174, a) will work because method dispatching (mostly) works
on the first argument to the function.  That is, plot(1:174, <whatever>)
does a default plot because 1:174 has no class, but plot(<something of class
"data.frame">, <whatever>) does a special plot designed for data frames.
Here's a simple toy example, ...

    > x <- 1:10
    > class(x) <- "test"
    > plot.test <- function(x, y) { cat("A custom plot") }
    > plot(1:10, x)  # produces a default scatterplot
    > plot(x, 1:10)
    A custom plot

Hope that helps.


r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list