Hi all, I have a bit of a newbie question here that I hope y'all can help with. I've run a cluster analysis using hclust on about 500 objects (using R1.3 under Win 2000). The problem is that the tips of the dendrogram are so close together on the plot that the labels overlap and are unreadable. I've used "cex" to reduce the label sizes but this isn't sufficient with so many objects on the plot. I've looked at "cutree" but this seems to show a summary of the upper levels of the dendrogram (?) and I want to see the details at the bottom. Basically I just want to print the dendrogram over several pages so that the plot is stetched out horizontally and the tip labels become readable. Is that possible? More generally, how does one print out large plots over several pages?? Can that be done? Many thanks for any help, Duncan ******************************************* Dr. Duncan Mackay Biology Flinders University GPO Box 2100 Adelaide S.A. 5001 AUSTRALIA Ph (08) 8201 2627 FAX (08) 8201 3015 http://www.bio.flinders.edu.au/dam/damres.htm -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- 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 _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
fharrell@virginia.edu
2001-Aug-28 02:01 UTC
[R] plotting dendrograms from cluster analyses
This may not be a solution to your problem but I think it would be worthwhile to develop a dendogram graphics driver that uses AT&T's graphviz package, which has its only command language and produces very detailed directed and undirected graphs. You can find graphviz at http://www.research.att.com/sw/tools/graphviz The manuals for the programs that make up graphviz have many nice examples. Frank Harrell Duncan Mackay wrote:> > Hi all, > I have a bit of a newbie question here that I hope y'all can help with. I've > run a cluster analysis using hclust on about 500 objects (using R1.3 under > Win 2000). The problem is that the tips of the dendrogram are so close > together on the plot that the labels overlap and are unreadable. I've used > "cex" to reduce the label sizes but this isn't sufficient with so many > objects on the plot. I've looked at "cutree" but this seems to show a > summary of the upper levels of the dendrogram (?) and I want to see the > details at the bottom. Basically I just want to print the dendrogram over > several pages so that the plot is stetched out horizontally and the tip > labels become readable. Is that possible? More generally, how does one print > out large plots over several pages?? Can that be done? > Many thanks for any help, > Duncan > > ******************************************* > Dr. Duncan Mackay > Biology > Flinders University > GPO Box 2100 > Adelaide > S.A. 5001 > AUSTRALIA > > Ph (08) 8201 2627 FAX (08) 8201 3015 > > http://www.bio.flinders.edu.au/dam/damres.htm > > -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- > 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 > _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._-- Frank E Harrell Jr Prof. of Biostatistics & Statistics Div. of Biostatistics & Epidem. Dept. of Health Evaluation Sciences U. Virginia School of Medicine http://hesweb1.med.virginia.edu/biostat -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- 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 _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Under Windows 2000 (since you are using this environment), some printer
drivers have options for "poster printing". For instance, with the
later
driver of the HP Deskjet 990cxi, you can print in up to 4x4 sheets of paper.
This could be an easy solution to your problem. Just select the desired
options in the printer driver. If your printer driver doesn't exhibit such
features, you still can save a metafile version of your graph and import it
in a vector drawing program that offer this possibility, like Corel Draw (I
think this option exists since version 5 or 6). I don't believe this should
be a feature included in R. It is really a printer driver job!
Also, (I haven't looked at it) but there is a package "pinktoe" in
CRAN that
convert a tree into a serie of html objects you can transverse interactively
in a web browser. This could be another convenient solution to explore your
very large tree.
Best regards,
Philippe Grosjean
..........]<(({?<...............<?}))><...............................
) ) ) ) ) __ __
( ( ( ( ( |__) | _
) ) ) ) ) | hilippe |__)rosjean
( ( ( ( ( Marine Biol. Lab., ULB, Belgium
) ) ) ) ) __
( ( ( ( ( |\ /| |__)
) ) ) ) ) | \/ |ariculture & |__)iostatistics
( ( ( ( (
) ) ) ) ) e-mail: phgrosje at ulb.ac.be or phgrosjean at sciviews.org
( ( ( ( ( SciViews project coordinator (http://www.sciviews.org)
) ) ) ) ) tel: 00-32-2-650.29.70 (lab), 00-32-2-673.31.33 (home)
( ( ( ( (
) ) ) ) ) "I'm 100% confident that p is between 0 and 1"
( ( ( ( ( L. Gonick & W. Smith (1993)
) ) ) ) )
.......................................................................
>Hi all,
>I have a bit of a newbie question here that I hope y'all can help with.
I've>run a cluster analysis using hclust on about 500 objects (using R1.3 under
>Win 2000). The problem is that the tips of the dendrogram are so close
>together on the plot that the labels overlap and are unreadable. I've
used
>"cex" to reduce the label sizes but this isn't sufficient with
so many
>objects on the plot. I've looked at "cutree" but this seems to
show a
>summary of the upper levels of the dendrogram (?) and I want to see the
>details at the bottom. Basically I just want to print the dendrogram over
>several pages so that the plot is stetched out horizontally and the tip
>labels become readable. Is that possible? More generally, how does one
print>out large plots over several pages?? Can that be done?
>Many thanks for any help,
>Duncan
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
> Hi all, > I have a bit of a newbie question here that I hope y'all can help with.I've> run a cluster analysis using hclust on about 500 objects (using R1.3 under > Win 2000). The problem is that the tips of the dendrogram are so close > together on the plot that the labels overlap and are unreadable. I've used > "cex" to reduce the label sizes but this isn't sufficient with so many > objects on the plot. I've looked at "cutree" but this seems to show a > summary of the upper levels of the dendrogram (?) and I want to see the > details at the bottom. Basically I just want to print the dendrogram over > several pages so that the plot is stetched out horizontally and the tip > labels become readable. Is that possible? More generally, how does oneprint> out large plots over several pages?? Can that be done? > Many thanks for any help, > DuncanI have done clusterings with ca. 300 items. For those, I found that a paper size of 25 by 25 inches was more than enough.I don't remember, but I think I set the cex to something like 0.5~0.7 on those. When printed out on multiple pages the trees were perfectly readable. Regards, Kuba Fast -- OnetKomunikator - porozumiesz sie z innymi [ http://ok.onet.pl/instaluj.html ] -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- 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 _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
>In university I learned S-Plus, which has a special function exactly for >this problem called subtree, where you can plot a region of a whole tree. >Now in R I missed that function several times. Is there a reason for not >including it in R?>ThomasExact. I've looked in Splus and found the nice subtree command. In R v. 1.3.0 and all packages installed, I've found only snip.rpart for rpart trees. Thus it seems a subtree function should be a good addition to R. I am doing a package in which this could be added (for a very preliminary version, in French, look at http://www.sciviews.org/_passtec), though it is not a top priority for me now. If someone else want to do it, just notify me... otherwise, I'll do it in a few months. Best regards, Philippe Grosjean ...........]<(({?<...............<?}))><............................... ) ) ) ) ) __ __ ( ( ( ( ( |__) | _ ) ) ) ) ) | hilippe |__)rosjean ( ( ( ( ( Marine Biol. Lab., ULB, Belgium ) ) ) ) ) __ ( ( ( ( ( |\ /| |__) ) ) ) ) ) | \/ |ariculture & |__)iostatistics ( ( ( ( ( ) ) ) ) ) e-mail: phgrosje at ulb.ac.be or phgrosjean at sciviews.org ( ( ( ( ( SciViews project coordinator (http://www.sciviews.org) ) ) ) ) ) tel: 00-32-2-650.29.70 (lab), 00-32-2-673.31.33 (home) ( ( ( ( ( ) ) ) ) ) "I'm 100% confident that p is between 0 and 1" ( ( ( ( ( L. Gonick & W. Smith (1993) ) ) ) ) ) ....................................................................... -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- 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 _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._