Re: [Rd] View() + "End" key on Ubuntu=segfault
Ben Bolker wrote:
>I can repeatably crash R (segfault)
>
> by doing
>
> n <- 10
> z <- data.frame(a=1:n,b=1:n)
> View(z)
>
> and then hitting the "End" key on my keyboard.
>
>I haven't got debugging going yet, but running under
> gdb (without debugging symbols) does give this:
>
> 0xb7b63583 in strlen () from /lib/tls/i686/cmov/libc.so.6
>
AArrgh There's always one more, isn't there?
Same thing with Fedora, and you don't even have to create the data
frame: View(airquality), press End suffices. I have
(gdb) bt
#0 0x009c2153 in strlen () from /lib/libc.so.6
#1 0x001d9289 in drawwindow (DE=0x96ca3e8)
at ../../../../R/src/modules/X11/dataentry.c:1380
#2 0x001da0ec in doSpreadKey (DE=0x96ca3e8, key=,
event=)
at ../../../../R/src/modules/X11/dataentry.c:1701
#3 0x001dc93a in R_ProcessX11Events (data=0x0)
at ../../../../R/src/modules/X11/dataentry.c:1380
1380printstring(DE, p, strlen(p), 0, i - DE->colmin + 1, 0);
(gdb) p p
No symbol "p" in current context.
(gdb) l
1375const char *p;
1376int i;
1377
1378for (i = DE->colmin; i <= DE->colmax; i++) {
1379p = get_col_name(DE, i);
1380printstring(DE, p, strlen(p), 0, i - DE->colmin + 1, 0);
1381}
1382for (i = DE->rowmin; i <= DE->rowmax; i++) {
1383sprintf(clab, DE->labform, i);
1384printstring(DE, clab, strlen(clab), i - DE->rowmin + 1,
0, 0);
at which point the optimizer seems to have messed things up too much for
further tracing.
> R version 2.6.2 (2008-02-08)
> i486-pc-linux-gnu
>
> [Ubuntu Gutsy]
>
> locale:
> LC_CTYPE=en_US.UTF-8;LC_NUMERIC=C;LC_TIME=en_US.UTF-8;LC_COLLATE=en_US.UTF-8;LC_MONETARY=en_US.UTF-8;LC_MESSAGES=en_US.UTF-8;LC_PAPER=en_US.UTF-8;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=en_US.UTF-8;LC_IDENTIFICATION=C
>
> attached base packages:
> [1] stats graphics grDevices utils datasets methods base
>
> __
> R-devel@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>
--
O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B
c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K
(*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918
~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] View() + "End" key on Ubuntu=segfault
It's a generic problem with the X11 dataeditor using End. It was trying to start at column 0. Fixed now. On Sat, 9 Feb 2008, Ben Bolker wrote: > > I can repeatably crash R (segfault) > > by doing > > n <- 10 > z <- data.frame(a=1:n,b=1:n) > View(z) > > and then hitting the "End" key on my keyboard. > > I haven't got debugging going yet, but running under > gdb (without debugging symbols) does give this: > > 0xb7b63583 in strlen () from /lib/tls/i686/cmov/libc.so.6 > > R version 2.6.2 (2008-02-08) > i486-pc-linux-gnu > > [Ubuntu Gutsy] > > locale: > LC_CTYPE=en_US.UTF-8;LC_NUMERIC=C;LC_TIME=en_US.UTF-8;LC_COLLATE=en_US.UTF-8;LC_MONETARY=en_US.UTF-8;LC_MESSAGES=en_US.UTF-8;LC_PAPER=en_US.UTF-8;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=en_US.UTF-8;LC_IDENTIFICATION=C > > attached base packages: > [1] stats graphics grDevices utils datasets methods base > > __ > R-devel@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-devel > -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] print.fitdistr buglet
On 09/02/2008 11:33 AM, Enrico Rossi wrote: > Dear developers, > > There's a small bug in print.fitdistr that can cause output to be printed > twice, but only if print is called explicitly: > >> fit<-fitdistr(rt(1000,3),"t") > There were 11 warnings (use warnings() to see them) >> fit > m sdf > -0.021817231.001452963.13723878 > ( 0.03865057) ( 0.03999447) ( 0.33298377) >> print(fit) > m sdf > -0.021817231.001452963.13723878 > ( 0.03865057) ( 0.03999447) ( 0.33298377) > m sdf > -0.021817231.001452963.13723878 > ( 0.03865057) ( 0.03999447) ( 0.33298377) > > Most likely, the function print.fitdistr doesn't return the input invisibly. I've fixed a number of instances of this kind of thing in the base packages now. A related kind of error I also fixed were print methods that returned an invisible NULL, or something other than what was printed. I didn't touch the recommended packages. These fixes are in R-patched now. Duncan Murdoch __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] help with R rendering engine
Hi Michael Lawrence wrote: > On Feb 7, 2008 4:49 PM, Mark W Kimpel <[EMAIL PROTECTED]> wrote: > >> I'm doing some work on a potential patch to the Bioconductor package >> Rgraphviz and have some questions on code that is contained in engine.c. >> In particular, I am developing some custom shapes using polygon and need >> to make sure that, with rendering, the line connecting the centers of >> two polygons stops at the border of each polygon. The polygons can be >> transparent, so the option of just rendering the lines first won't work. >> >> Is there some help for me with an R internal or perhaps a higher level >> function? >> > > I have to admit I am not that familiar with R graphics, but a very common > way to achieve this in general is to use clipping. That is, you would set > the regions covered by your node polygons as the "clip" and then draw your > edges. Anything drawn onto the clip region is discarded. You would then > reset the clip and continue drawing. > > I'm pretty sure that base R does not support the complex clip regions you > would require. However, it would be possible to do this in R by rendering to That's right. R's graphics engine only copes with rectangular clipping regions. OTOH, the 'grid' package does have a facility to determine locations on the edges of objects (see grobX() abnd grobY()). This is not fully-featured, but for convex polygons it might give you what you want. There's an example or two in http://www.stat.auckland.ac.nz/~paul/Talks/rgraphs.pdf Paul > pixmaps in memory and then compositing the pixmaps. The first part is > supported by the cairoDevice package and the latter part by RGtk2 (in > particular its binding to GDK). I'm sure you'd rather not introduce such a > dependency on Rgraphviz, but it's an option. I think the graphviz developers > are moving towards GTK+/Cairo in terms of plugins and other work. > > I checked libgd (already a dependency of graphviz) but found that it only > supports a single rectangle for clipping. > > Anyway, I hope this helps a little and good luck, > Michael > > >> Thanks, >> Mark >> -- >> >> Mark W. Kimpel MD ** Neuroinformatics ** Dept. of Psychiatry >> Indiana University School of Medicine >> >> 15032 Hunter Court, Westfield, IN 46074 >> >> (317) 490-5129 Work, & Mobile & VoiceMail >> (317) 204-4202 Home (no voice mail please) >> >> mwkimpelgmailcom >> >> __ >> R-devel@r-project.org mailing list >> https://stat.ethz.ch/mailman/listinfo/r-devel >> > > [[alternative HTML version deleted]] > > __ > R-devel@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-devel -- Dr Paul Murrell Department of Statistics The University of Auckland Private Bag 92019 Auckland New Zealand 64 9 3737599 x85392 [EMAIL PROTECTED] http://www.stat.auckland.ac.nz/~paul/ __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] help with R rendering engine
Paul and Michael, Thanks for the information. Yes, it appears that what I want is to be able to present R with a convex polygon for the clipping path. As I waited for comments (for some reason Michael's email did not get to me), I began working on my own methods. I am essentially using a matrix in R to represent pixel-space and using my own edge-detection algorithm. As I write this, I believe I am almost there, but testing awaits tomorrow. The real question will be how fast it actually runs. I'll report back later this week. Mark Mark W. Kimpel MD ** Neuroinformatics ** Dept. of Psychiatry Indiana University School of Medicine 15032 Hunter Court, Westfield, IN 46074 (317) 490-5129 Work, & Mobile & VoiceMail (317) 204-4202 Home (no voice mail please) mwkimpelgmailcom ** Paul Murrell wrote: > Hi > > > Michael Lawrence wrote: >> On Feb 7, 2008 4:49 PM, Mark W Kimpel <[EMAIL PROTECTED]> wrote: >> >>> I'm doing some work on a potential patch to the Bioconductor package >>> Rgraphviz and have some questions on code that is contained in engine.c. >>> In particular, I am developing some custom shapes using polygon and need >>> to make sure that, with rendering, the line connecting the centers of >>> two polygons stops at the border of each polygon. The polygons can be >>> transparent, so the option of just rendering the lines first won't work. >>> >>> Is there some help for me with an R internal or perhaps a higher level >>> function? >>> >> I have to admit I am not that familiar with R graphics, but a very common >> way to achieve this in general is to use clipping. That is, you would set >> the regions covered by your node polygons as the "clip" and then draw your >> edges. Anything drawn onto the clip region is discarded. You would then >> reset the clip and continue drawing. >> >> I'm pretty sure that base R does not support the complex clip regions you >> would require. However, it would be possible to do this in R by rendering to > > > That's right. R's graphics engine only copes with rectangular clipping > regions. > > OTOH, the 'grid' package does have a facility to determine locations on > the edges of objects (see grobX() abnd grobY()). This is not > fully-featured, but for convex polygons it might give you what you want. > There's an example or two in > http://www.stat.auckland.ac.nz/~paul/Talks/rgraphs.pdf > > Paul > > >> pixmaps in memory and then compositing the pixmaps. The first part is >> supported by the cairoDevice package and the latter part by RGtk2 (in >> particular its binding to GDK). I'm sure you'd rather not introduce such a >> dependency on Rgraphviz, but it's an option. I think the graphviz developers >> are moving towards GTK+/Cairo in terms of plugins and other work. >> >> I checked libgd (already a dependency of graphviz) but found that it only >> supports a single rectangle for clipping. >> >> Anyway, I hope this helps a little and good luck, >> Michael >> >> >>> Thanks, >>> Mark >>> -- >>> >>> Mark W. Kimpel MD ** Neuroinformatics ** Dept. of Psychiatry >>> Indiana University School of Medicine >>> >>> 15032 Hunter Court, Westfield, IN 46074 >>> >>> (317) 490-5129 Work, & Mobile & VoiceMail >>> (317) 204-4202 Home (no voice mail please) >>> >>> mwkimpelgmailcom >>> >>> __ >>> R-devel@r-project.org mailing list >>> https://stat.ethz.ch/mailman/listinfo/r-devel >>> >> [[alternative HTML version deleted]] >> >> __ >> R-devel@r-project.org mailing list >> https://stat.ethz.ch/mailman/listinfo/r-devel > __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
