rkward does not include the stats package

Bug #224480 reported by JSmith
8
This bug affects 1 person
Affects Status Importance Assigned to Milestone
r-base (Ubuntu)
Invalid
Undecided
Unassigned

Bug Description

Binary package hint: rkward

I am and using 8.04.

I am using the most recent version of rkward from the deb http://cran.mtu.edu/bin/linux/ubuntu hardy/ repository.

I expected to be able to run ANOVAs and the like.

When I attempt to run ANOVAs, MANOVAs, etc., rkward tells me that I need the stats package because it was not included in the install. This would seem to be a rather significant omission.

Revision history for this message
Thomas Friedrichsmeier (tfry) wrote :

RKWard relies on the regular R installation. I'm not quite sure, whether you use the regular ubuntu R packages. If you do not, please report this, whereever you got your R packages from. If you are using ubuntu R packages, you might want to check, whether you have r-recommended installed (which should generally be pulled in automatically, IIRC, but maybe this is the problem).
Either way, I'll move this report to r-base for the R maintainers to look at.

Revision history for this message
Cesare Tirabassi (norsetto) wrote :

What stats package are you talking about? Can you be more specific about the binary (or source) name?
Also, there is no rkward (or at least I can't find it) in the cran repo.
Can you try to to remove the packages you have installed from the cran repo and install the ones from our repo? This should at least tell us if there is something we ought to do or if it is a problem with a package from the cran repo.

Changed in r-base:
status: New → Incomplete
Revision history for this message
Thomas Friedrichsmeier (tfry) wrote :

I'll fill in some blanks:

The "stats" package is a part of a basic R installation. It seems to be included in r-base-core (so, please ignore my comment about r-recommended). In R terminology, "stats" is a "base package", so it would not generally be fetched from CRAN, manually. Since I don't use Ubuntu myself, I can't tell, whether or not there may be a problem with r-base-core in ubuntu. In debian/unstable it seems to work fine. You can test by calling

> library ("stats")
or
> example (summary.manova)

RKWard is a GUI frontend for R. This is not packaged in the form of an R package. Rather it is an application linked against R (and depends on r-base-core). RKWard is included in ubuntu. RKWard offers R package management facilities, but really just uses the regular R mechanisms, so I can't see, how this could be specific to RKWard (and again: no problem, here), which is why I moved the report to r-base.

From the bug report it appears that something is broken in the user's R installation. I can't tell, whether this is a problem in ubuntu packages.

@JSmith: Please do provide the additional information requested by Cesare, though. Also information on your installed versions of r-base, r-base-core, and rkward might help.

Revision history for this message
Cesare Tirabassi (norsetto) wrote :

Thanks for the additional info Thomas, its much clearer now. As part of the gfortran transition we have recently transitioned to r-base-core 2.6.2. rkward has not been rebuilt with this new version of r-base, could this be at the root of this problem? I don't see any rkward package being in the cran repo, so I suspect it could be a problem on our side (but I need to see what JSmith has installed to confirm or otherwise).

@JSmith: can you tell us what packages you have installed? You can use dpkg or apt-cache (for instance dpkg -l or apt-cache policy).

Revision history for this message
Cesare Tirabassi (norsetto) wrote :

In case it can be useful, this is what ldd is telling me about rkward 0.4.9:

root@norsetto:/# ldd /usr/lib/R/site-library/rkward/libs/rkward.so
        libR.so => /usr/lib/R/lib/libR.so (0x00002b42ef56c000)
        libc.so.6 => /lib/libc.so.6 (0x00002b42efa77000)
        libblas.so.3gf => /usr/lib/libblas.so.3gf (0x00002b42efdda000)
        libgfortran.so.2 => /usr/lib/libgfortran.so.2 (0x00002b42f0054000)
        libm.so.6 => /lib/libm.so.6 (0x00002b42f0313000)
        libgcc_s.so.1 => /lib/libgcc_s.so.1 (0x00002b42f0595000)
        libreadline.so.5 => /lib/libreadline.so.5 (0x00002b42f07a3000)
        libpcre.so.3 => /usr/lib/libpcre.so.3 (0x00002b42f09e3000)
        libbz2.so.1.0 => /lib/libbz2.so.1.0 (0x00002b42f0c0a000)
        libz.so.1 => /usr/lib/libz.so.1 (0x00002b42f0e1a000)
        libdl.so.2 => /lib/libdl.so.2 (0x00002b42f1031000)
        /lib64/ld-linux-x86-64.so.2 (0x0000555555554000)
        libncurses.so.5 => /lib/libncurses.so.5 (0x00002b42f1236000)

Revision history for this message
Thomas Friedrichsmeier (tfry) wrote :

> rkward has not been rebuilt with this new version of r-base, could this be at the root of this problem?

I *think* rkward will not start at all unless it is rebuilt (see bug #221399; libRblas.so is no longer included in r-base-core 2.6.2), so this should be a different issue. On the other hand, I can't rule out the possibility that this is another symptom of the same underlying issue.

Well, it's all just speculation without knowing the exact installed versions, and more so, since I don't have a ubuntu system to test with. If it does not pull too many dependencies for a quick test, maybe you could install rkward and see what happens (there is an integrated "R Console", where you can try the example lines - if it starts at all).

Revision history for this message
Thomas Friedrichsmeier (tfry) wrote :

> ldd /usr/lib/R/site-library/rkward/libs/rkward.so

Thanks. This is only a mostly empty "dummy" library, though. The main binary is /usr/bin/rkward.bin. Just type rkward on a regular console to (try to) start it (chances are, it won't work in your installation due to bug #221399).

Revision history for this message
Cesare Tirabassi (norsetto) wrote :

As you can see I'm running the summary.manova example without problems with the ubuntu rebuilt version?

Revision history for this message
JSmith (smith-js1) wrote :

It appeared to be a mirror issue. I reinstalled from another mirror, and it works fine. Thanks!

Revision history for this message
Cesare Tirabassi (norsetto) wrote :

Thanks to you for reporting, marking this as invalid then.

Changed in r-base:
status: Incomplete → Invalid
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.