R to automate scatter plots

Michael,

Thanks for all your help.

I have things working pretty nicely now and will post back to my thread
incase
any one else needs a solution or has something to add but before I do I want
to
understand how to print all plots to one pdf.

So far this is not working for me. If I make a new pdf("all.pdf")
after the for loop, and then repeat the plot command, followed by a
dev.off(),
ie
for () {
# ?all the individual pdfs here
}
pdf("allpdf")
plot()
devoff.()

I only get one page on the all.pdf with the plot of the last data object, so
again it is writing over the last pdf

Cynthia


On Nov 10, 2011, at 11:48 AM, R. Michael Weylandt wrote:

It's usually encouraged to reply to the list as well for
threading/archive reasons. (And because occasionally your first
respondent will say something wrong and require correction -- I was
called out just this morning...oops)

Yes, you seem to be using it right to add a title to the plot but you
have a few syntactical errors (one of which I think came from my first
reply): I think you would prefer this:

paste(data$fpkma, data$pdkmb, main =
paste(substr(files[i],1,nchar(files[i])-4), ".pdf", collapse = ""),
xlab = "FPKM First", ylab = "FPKM Second")

The placement of pdf() is irrelevant to how you get the names: whether
you put it outside or inside the loop, you can still either name it
directly by adding an argument to pdf() or use the default behavior:
the placement only affects whether you want one pdf with all your
plots or multiple pdfs of one plot each. The only other thing is that
you should call pdf() before plot() to make sure the pdf is ready to
take the plot.

Michael

On Thu, Nov 10, 2011 at 11:20 AM, Cynthia Lee Page
<Cynthia.Page at colorado.edu> wrote:

HI Michael,

You are exactly right about the object data. In my tweaking I messed up
all
my .csv and they were no longer .csv's
So there was ?no data object.

I will try what you suggested regarding generating the pdf files and
naming
them. I think I can use the same technique to
add a title to each plot I am not sure if I can use paste(
substr(files[i],1, nchar(files[i]-4)), ".pdf" below in the plot title as:

plot(data$fpkma, data$fpkmb, main=paste( substr(files[i],1,
nchar(files[i]-4)), ".pdf", xlab = "FPKM First", ylab="FPKM Second")

but I will try it.

?If I put the pdf() outside the loop the plots may go directly to
Rplot.pdf.
I think I have to call pdf() before calling plot() inorder to customize
the
name.

Now I am ramblin!

One other thing. I have been responding via reply to sender directly to
those who have responded to me ?- I have not posted ?to the list.

Should I be?

Thanks so much,

Cynthia



On Nov 9, 2011, at 6:28 PM, R. Michael Weylandt wrote:

It sounds like something is going wrong in your data read -- the
warnings indicate that R probably isn't data to the plotting commands.
My totally off the wall guess is that if your data is coming by way of
excel, the commas are leading to your data becoming characters and
hence not plotting nicely, but that's just a guess. add browser()
inside your loop between the read commands and the plot just to see
what your data actually looks like inside R and try to track down the
error.

As to your pdf() question -- you can also just put it outside the loop
and close dev.off() after the loop. Then all your plots will be in one
pdf. Though, perhaps this is what you were looking for as far as name
manipulation:

pdf(file = paste( substr(files[i],1, nchar(files[i]-4)), ".pdf", sep =
"")

Michael

On Wed, Nov 9, 2011 at 5:32 PM, David Winsemius <dwinsemius at comcast.net>
wrote:

On Nov 9, 2011, at 4:22 PM, Cynthia Lee Page wrote:

Hi R people!

I have a directory of .csv files I would like to make into objects
then
scatter plots. I have been having varying degrees of progress. I was
able
make an object of all files, loop through it, and make a pdf of the
last
file I looped through. I kept renaming the pdf so instead of ending up
with
27 pdfs I got one, with the data from the last file

I have been tweaking with it and now can't even make the data object
and
I
am not sure why.

I am a bit brain dead at this point :)

I am new to R and have been programming in perl - but not all that
long

Could you please have ?al look at it..

here is the script I have been using

# source of this code below
#http://cran.r-project.org/doc/contrib/Lemon-kickstart/kr_scrpt.html

# store the current directory
initial.dir<-getwd()
# change to the new directory


setwd("/data/homes/ccpage/ngs/Argueso/Tophat/flocculated/cuffdiff/fpkmgt")

# source of this code below
# https://stat.ethz.ch/pipermail/r-help/2008-March/158336.html


files <- Sys.glob("*.csv") ?# get names of files to process
#result <- numeric(length(files)) ?# preallocate assuming single value
from each file

for (i in seq_along(files)){
# want to give each object a unique name would like to use file[i]
MINUS
the .csv extention regex
#test<-files[i] # tried to use as variable to name each pdf this
object
is
the name of last file in loop

?data <- read.csv(files[i])

# I want to name the pdf the same name as the object with a .pdf
extention
here I think it will be file[i].csv.pdf
# I don't know how to use regex in R I could
readLines(objectnames.txt)
and loop through those as well

?pdf("data.pdf")

At this point you might have been better off if you had just typed:

? pdf()

The default name for a pdf document is set by this code from the help
page
for pdf()
pdf(file = ifelse(onefile, "Rplots.pdf", "Rplot%03d.pdf"),
Notice that "%03d". That means the system pots in a number tthat is one
grater than the largest current Rplot_N.pdf in the directory.

?plot(data$fpkma,data$fpkmb, main="Scatter plot of data",xlab="FPKM of
First Time Point",ylab="FPKM of Second Time Point")
?dev.off()
}

# change back to the original directory
setwd(initial.dir)
############################################################

the command I have been using :
R CMD BATCH /data/homes/ccpage/ngs/rscripts/test_for.R

The Rout

# source of this code below
#http://cran.r-project.org/doc/contrib/Lemon-kickstart/kr_scrpt.html

# store the current directory
initial.dir<-getwd()
# change to the new directory



setwd("/data/homes/ccpage/ngs/Argueso/Tophat/flocculated/cuffdiff/fpkmgt")

# source of this code below
# https://stat.ethz.ch/pipermail/r-help/2008-March/158336.html

files <- Sys.glob("*.csv") ?# get names of files to process
#result <- numeric(length(files)) ?# preallocate assuming single
value
from each file

for (i in seq_along(files)){

+ # want to give each object a unique name would like to use file[i]
MINUS
the .csv extention regex
+ #test<-files[i] # tried to use as variable to name each pdf this
object
is the name of last file
+
+ ? ?data <- read.csv(files[i])
+
+ # I want to name the pdf the same name as the object with a .pdf
extention here I think it will be file[i].csv.pdf
+ # I don't know how to use regex in R I could
readLines(objectnames.txt)
and loop through those as well
+
+ ? ? pdf("data.pdf")
+ ? ? plot(data$fpkma,data$fpkmb,main="Scatter plot of
data",xlab="FPKM
of
First Time Point",ylab="FPKM of Second Time Point")
+ ? ? dev.off()
+ }
Error in plot.window(...) : need finite 'xlim' values

Without the data that created that error, we are not going to be able
to
give a clear answer.

Calls: plot -> plot.default -> localWindow -> plot.window
In addition: Warning messages:
1: In min(x) : no non-missing arguments to min; returning Inf
2: In max(x) : no non-missing arguments to max; returning -Inf
3: In min(x) : no non-missing arguments to min; returning Inf
4: In max(x) : no non-missing arguments to max; returning -Inf
Execution halted

Thanks for any help!\

David Winsemius, MD
West Hartford, CT