http://blog.gmane.org/gmane.linux.gentoo.science hourly 1 1901-01-01T00:00+00:00 http://gmane.org/img/gmane-25t.png http://gmane.org http://permalink.gmane.org/gmane.linux.gentoo.science/1965 <pre>Hi Christoph, thanks for the detailed answer. I have made a few changes to the ebuild, implementing what you suggested. The only thing missing right now is a better way to deal with the Makefile. I am working on that. In the meantime I have created a pull request. Thanks again, nick On Sun, May 12, 2013 at 5:08 PM, Christoph Junghans &lt;ottxor&lt; at &gt;gentoo.org&gt;wrote: </pre> Nicolas Bock 2013-05-13T16:22:20 http://permalink.gmane.org/gmane.linux.gentoo.science/1964 <pre>Hi Nick, thanks for your contribution. I think there is no general answer to your question. It depends if it is a disadvantage of running the mpi-version with 1 MPI process over running the serial version. I imagine that LAMMPS enable some special optimization if mpi is disabled as it can assume that everything happens locally and no mpi message need to be send. On the other hand, if the build of LAMMPS takes hours, one might prefer building it only once. Do LAMMPS offer other palatalizations like OpenMP or Threads to use multiple cores? From a brief look at our patch I see the following issues: - I am not sold on this Makefile patch and patching of the patch approach, seems a bit fragile. It might be better to use the default gcc Makefile and patch that. I prefer adding CC,CFLAGS etc, on emake comandline, you could create a function to avoid writing it on every emake call. (my_emake() { emake "$&lt; at &gt;" CC= CFLAGS=; } ) - sed calls need to secured with "|| die" - DEPEND should be static, DEPEND="mpi? ( vitrual/mpi )" should do it - virtual/fortran is missing in DEPEND or use fortran-2.eclass and FORTRAN_NEEDED="fortran" instead. - lammps-gzip enables normal gzip support, so the use flags should be called gzip. If you are unsure about some technical details in an ebuild, create a pull request on github and somebody will give you advice in the review. Otherwise good job, looking forward to more commits from you, Christoph 2013/5/12 Nicolas Bock &lt;nicolasbock&lt; at &gt;gmail.com&gt;: </pre> Christoph Junghans 2013-05-12T23:08:26 http://permalink.gmane.org/gmane.linux.gentoo.science/1963 <pre>Hi, I recently added a lammps ebuild to the science overlay. It so far only builds the serial version, and has very limited support for build time options. I now added the mpi use flag and changed the ebuild so that it builds the mpi version. I have attached a patch against master that does this. I would like to get some feedback on those changes. Are they acceptable in this form? What should I have done differently? How would I go about writing an ebuild that builds both, the serial _and_ the mpi version in case the mpi use flag is set? Thanks already, nick </pre> Nicolas Bock 2013-05-12T21:22:39 http://permalink.gmane.org/gmane.linux.gentoo.science/1962 <pre>wow, thanks. now it works! (I had to recompile maxima too, even though only ecls had the wrong version.) thanks again for the quick solution. </pre> Piotr Sawuk 2013-05-05T18:15:38 http://permalink.gmane.org/gmane.linux.gentoo.science/1961 <pre> Yes, it does. I've got the same as you. Reinstalling both deps should fix the problem: emerge -1 =dev-lisp/ecls-12.12.1 emerge -1 =sci-mathematics/maxima-5.29.1-r1 </pre> Christopher Schwan 2013-05-05T08:56:25 http://permalink.gmane.org/gmane.linux.gentoo.science/1960 <pre> Hi, Can you make sure your ecls is from the sage-on-gentoo overlay? The bits of maxima required for your example to work can only be built with ecls from the overlay. In doubt do the following: emerge -1v =ecls-12.12.1::sage-on-gentoo followed by emerge -1v maxima::sage-on-gentoo Francois </pre> Francois Bissey 2013-05-05T08:55:31 http://permalink.gmane.org/gmane.linux.gentoo.science/1959 <pre> x=var('x');simplify(x+1-x) and as a result I got: RuntimeError: ECL says: Module error: Don't know how to REQUIRE MAXIMA. if anyone else has sage installed, does above work? if yes, what are the versions of software sage is depending on? "equery --no-color g =sci-mathematics/sage-5.8-r3" shows me: sci-mathematics/sage-5.8-r3: [ 0] sci-mathematics/sage-5.8-r3 [ 1] dev-libs/gmp-5.0.2_p1 [ 1] dev-libs/mpfr-3.1.1 [ 1] dev-libs/mpc-1.0.1 [ 1] dev-libs/ntl-5.5.2 [ 1] dev-libs/ppl-0.11.2-r1 [ 1] dev-lisp/ecls-12.12.1 [ 1] dev-python/numpy-1.7.1 [ 1] dev-python/cython-0.18-r1 [ 1] sci-mathematics/eclib-20120830 [ 1] sci-mathematics/gmp-ecm-6.4.4-r2 [ 1] sci-libs/flint-1.6 [ 1] sci-libs/fplll-3.0.12-r3 [ 1] sci-libs/givaro-3.7.1 [ 1] sci-libs/gsl-1.15 [ 1] sci-libs/iml-1.0.3-r2 [ 1] sci-libs/libcliquer-1.21_p0 [ 1] sci-libs/libgap-4.5.7_p2-r1 [ 1] sci-libs/linbox-1.3.2 [ 1] sci-libs/m4ri-20120613 [ 1] sci-libs/m4rie-20120613 [ 1] sci-libs/mpfi-1.5.1 [ 1] sci-libs/pynac-0.2.6-r1 [ 1] sci-libs/symmetrica-2.0-r2 [ 1] sci-libs/zn_poly-0.9-r3 [ 1] sci-mathematics/glpk-4.45 [ 1] sci-mathematics/lcalc-1.23-r4 [ 1] sci-mathematics/lrcalc-1.1.6_beta1 [ 1] sci-mathematics/pari-2.5.3-r2 [ 1] sci-mathematics/polybori-0.8.2-r4 [ 1] sci-mathematics/ratpoints-2.1.3-r3 [ 1] sci-mathematics/sage-baselayout-5.8-r1 [ 1] sci-mathematics/sage-clib-5.8 [ 1] sci-libs/libsingular-3.1.5-r2 [ 1] media-libs/gd-2.0.35-r3 [ 1] media-libs/libpng-1.5.15 [ 1] sys-libs/readline-6.2_p1 [ 1] sys-libs/zlib-1.2.7 [ 1] virtual/cblas-1.0 [ 1] dev-lang/python-2.7.3-r3 [ 1] dev-python/python-exec-0.3.1 [ 1] sci-mathematics/sage-notebook-0.10.4-r2 [ 1] sci-mathematics/sage-data-conway_polynomials-0.4-r1 [ 1] dev-lang/R-3.0.0 [ 1] dev-python/cvxopt-1.1.5-r2 [ 1] dev-python/gdmodule-0.56-r2 [ 1] dev-python/ipython-0.13.1 [ 1] dev-python/jinja-2.6-r1 [ 1] dev-python/matplotlib-1.1.1 [ 1] dev-python/mpmath-0.17 [ 1] dev-python/networkx-1.6 [ 1] dev-python/pexpect-2.0-r2 [ 1] dev-python/pycrypto-2.6-r2 [ 1] dev-python/rpy-2.3.6 [ 1] dev-python/sphinx-1.1.3-r6 [ 1] dev-python/sqlalchemy-0.8.1 [ 1] dev-python/sympy-0.7.2-r2 [ 1] media-gfx/tachyon-0.98.9-r2 [ 1] sci-libs/cddlib-094f-r2 [ 1] sci-libs/scipy-0.12.0 [ 1] sci-mathematics/flintqs-20070817_p8 [ 1] sci-mathematics/gap-4.5.7-r3 [ 1] sci-mathematics/genus2reduction-0.3_p8-r1 [ 1] sci-mathematics/gfan-0.5 [ 1] sci-mathematics/cu2-20060223 [ 1] sci-mathematics/cubex-20060128 [ 1] sci-mathematics/dikcube-20070912_p18 [ 1] sci-mathematics/maxima-5.29.1-r1 [ 1] sci-mathematics/mcube-20051209 [ 1] sci-mathematics/optimal-20040603 [ 1] sci-mathematics/palp-2.1 [ 1] sci-mathematics/sage-data-elliptic_curves-0.7 [ 1] sci-mathematics/sage-data-graphs-20120404_p4 [ 1] sci-mathematics/sage-data-polytopes_db-20100210_p2 [ 1] sci-mathematics/sage-doc-5.8-r2 [ 1] sci-mathematics/sage-extcode-5.8 [ 1] sci-mathematics/singular-3.1.5 [ 1] sci-mathematics/sympow-1.018.1_p11 [ 1] sys-libs/glibc-2.15-r3 [ 1] dev-tex/sage-latex-2.3.3_p2 [ 1] app-text/dvipng-1.14 [ 1] media-gfx/imagemagick-6.7.8.7 </pre> Piotr Sawuk 2013-05-05T08:34:00 http://permalink.gmane.org/gmane.linux.gentoo.science/1958 <pre> I see, I hadn't actually tested it with the latest ecls from the tree that copies an adsf file from adsf-2.33-r3. The change I mentioned works without that. Don't know what to do about that yet. By the way can you make ecls-12.12.1-r4 prefix friendly by replacing cp /usr/share/common-lisp/source/asdf/build/asdf.lisp with cp ${EPREFIX}/usr/share/common-lisp/source/asdf/build/asdf.lisp Francois </pre> Francois Bissey 2013-04-30T18:55:07 http://permalink.gmane.org/gmane.linux.gentoo.science/1957 <pre>How have you got so far? I get ;;; Loading #P"/usr/lib/ecl-12.12.1/asdf.fas" ;;; OPTIMIZE levels: Safety=2, Space=0, Speed=3, Debug=0 ;;; ;;; End of Pass 1.An error occurred during initialization: The function ASDF/INTERFACE::REMOVE-KEYS is undefined.. * ERROR: sci-mathematics/maxima-5.30.0 failed (install phase): * !!! newins: maxima.system.fasb does not exist It seems maxima-build.lisp (or some file included from it) is not compatible with the current asdf. Andrey </pre> grozin< at >gentoo.org 2013-04-30T16:13:34 http://permalink.gmane.org/gmane.linux.gentoo.science/1956 <pre>Quoting grozin&lt; at &gt;gentoo.org: Sorry to have taken so long to look at it. In maxima. The ecls library for maxima has changed name with the new asdf. So newins maxima.fasb maxima.fas should become: newins maxima.system.fasb maxima.fas Francois </pre> fbissey< at >slingshot.co.nz 2013-04-29T21:50:59 http://permalink.gmane.org/gmane.linux.gentoo.science/1954 <pre> There is https://bugs.gentoo.org/show_bug.cgi?id=465560 which I reported and seems to be related to slower hardware when running the self-tests. This problem is still present with sbcl-1.1.6-r4. I'm not sure about the threads issue that's also reported in the same bug. It seems to me that sbcl, at least here, isn't built in parallel. Steve </pre> Steven Trogdon 2013-04-22T03:26:24 http://permalink.gmane.org/gmane.linux.gentoo.science/1953 <pre>Hello *, Today I've committed to the tree the following packages (all pmasked): dev-lisp/asdf-2.33-r4 dev-lisp/uiop-2.33-r1 dev-lisp/sbcl-1.1.6-r4 dev-lisp/clisp-2.49-r7.ebuild dev-lisp/clozurecl-1.9_p1 dev-lisp/ecls-12.12.1-r4 dev-lisp/cmucl-20d-r3 sci-mathematics/maxima-5.30.0 sci-mathematics/fricas-1.2.0 The lisp stuff is from the lisp overlay. I haven't used the overlay specific eclasses, just the standard helpers (doins etc.). clozurecl-1.9-r3 from the overlay is a live ebuild; are some of the recent updates essential? I replaced it by the today's snapshot (I've put it to dev.gentoo.org). In clisp, I've corrected the dependency on pari (alas, it has to be &lt;pari-2.5), and added an upstream patch which may be needed for new gcc versions. All abuilds are EAPI=5. Now, when a lisp used for maxima or fricas is upgraded, these CASs will be upgraded automatically - this is exactly the main point of EAPI5. It would be great to unmask all these packages soon, not to wait for years. Any specific reasons not to do so for any of these packages? Should something be improved before unmasking? Please, test! And inform me. Testsuites of maxima and fricas are OK, so, the lisps cannot be broken. But I am not so sure about the asdf stuff. Andrey </pre> grozin< at >gentoo.org 2013-04-20T15:13:46 http://permalink.gmane.org/gmane.linux.gentoo.science/1952 <pre>justin vas escriure el dia dt, 09 abr 2013: This file generates data and samples (call it desc-000.bug) -----8&lt;--------------- data { Alpha &lt;- 3 Beta &lt;- 5 Sigma &lt;- 0.5 N &lt;- 500 for (i in 1:500) { epsilon[i] ~ dnorm(0, 1/Sigma^2) x[i] ~ dunif(0, 10) y[i] &lt;- Alpha + Beta*x[i] + epsilon[i] } } model { for (i in 1:500) { y[i] ~ dnorm(mu[i], tau) mu[i] &lt;- alpha + beta * x[i] } alpha ~ dnorm(0, 0.001) beta ~ dnorm(0, 0.001) tau &lt;- pow(sigma, -2) sigma ~ dgamma(0.001, 0.001) } -----8&lt;--------------- The sample script with 3 chains (m-000.jmd) that calls the previous file. -----8&lt;--------------- model in "desc-000.bug" compile, nchains(3) initialize update 10000 monitor alpha monitor beta monitor sigma update 10000 coda *, stem(coda_m-000-) parameters to "outs-000" exit -----8&lt;--------------- HTH, </pre> Xavier Fernández i Marín 2013-04-09T14:17:38 http://permalink.gmane.org/gmane.linux.gentoo.science/1951 <pre> Could you provide me a sample script which should run jags in parallel? Justin </pre> justin 2013-04-09T13:49:02 http://permalink.gmane.org/gmane.linux.gentoo.science/1950 <pre>Dear all, After some years with the official blas and lapack libraries from atlas several weeks ago these packages were discontinued in the official portage tree, being replaced by the science overlay. I've tried to get the same functionality than before, but with no luck. I used to run sci-mathematics/jags compiled against the atlas-threaded libraries, and when simulating more than one chain all processors were used in parallel. But now, with the current sci-libs/atlas libraries this is not the case anymore. I have also tried to link jags against openblas, but without luck as well. I have also tried to compile it against mkl using openmp and even compile jags without using the official ebuild. So my questions are: - Is there a way to check that atlas is using multithreaded capacities, besides checking the libraries towards which it is compiled? - What has atlas changed between the last unstable ebuild from the official portage tree and the last unstable ebuild from the science overlay? - Any guess on how I can make use of multithreading in jags in gentoo? Many thanks. </pre> Xavier Fernández i Marín 2013-04-09T13:40:07 http://permalink.gmane.org/gmane.linux.gentoo.science/1949 <pre> Hello! On Fri, 08 Mar 2013 20:53:07 +1300 Francois Bissey &lt;fbissey&lt; at &gt;slingshot.co.nz&gt; wrote: Yes, that indeed was the case. So I unmerged Sage and then merged the new one (5.7) which depends on sympy. So, now everything seems to work fine. Thanks! Vladimir ----- &lt;v_2e&lt; at &gt;ukr.net&gt; </pre> v_2e< at >ukr.net 2013-03-08T12:48:59 http://permalink.gmane.org/gmane.linux.gentoo.science/1948 <pre> That would be related to ipython from the log I can that this ipython interaction is coming through sage being called from mpmath itself called from sympy. Considering that sage 5.7 has switched to ipython 0.13.1 that your sage install is not completely consistent/complete. Do you have sage-baselayout at 5.7 and sage itself at 5.6? If that's the case one way out would be to unmerge sage, upgrade sympy, and then merge sage again. There are other ways but that's a bit more complicated. Francois </pre> Francois Bissey 2013-03-08T07:53:07 http://permalink.gmane.org/gmane.linux.gentoo.science/1947 <pre> Hello! On Fri, 08 Mar 2013 09:46:08 +1300 fbissey&lt; at &gt;slingshot.co.nz wrote: I uninstalled gmpy, but I still get an error: AttributeError: 'module' object has no attribute 'ipapi' Full log: http://bpaste.net/show/82199/ Vladimir ----- &lt;v_2e&lt; at &gt;ukr.net&gt; </pre> v_2e< at >ukr.net 2013-03-08T07:08:23 http://permalink.gmane.org/gmane.linux.gentoo.science/1946 <pre>Quoting v_2e&lt; at &gt;ukr.net: Sorry Vladimir, I got buried in stuff. Can you please uninstall gmpy? At least set MPMATH_NOGMPY=1 before emerging sympy. We should just ask everyone who uses sage-on-gentoo to mask gmpy. If you install sage, mpmath and therefore sympy won't work with gmpy. Don't even try. mpmath will install normally but not sympy because it import itself at compilation/install time. You can shutdown the use of gmpy in mpmath by setting MPMATH_NOGMPY. There is similarly a MPMATH_NOSAGE variable but it is broken in that one sage import is not under its control. Upstream mpmath knows about the last problem. Not sure when there will be a release to fix it. Francois </pre> fbissey< at >slingshot.co.nz 2013-03-07T20:46:08 http://permalink.gmane.org/gmane.linux.gentoo.science/1945 <pre> Hello! After re-emerging 'scons' and updating both portage and 'sage-on-gentoo' overlay, I do not observe build problems with sage-clib any more, but now sympy fails to build. Here is the build log: http://bpaste.net/show/81612/ Regards, Vladimir ----- &lt;v_2e&lt; at &gt;ukr.net&gt; </pre> v_2e< at >ukr.net 2013-03-05T16:09:20 http://permalink.gmane.org/gmane.linux.gentoo.science/1944 <pre>At Tue, 05 Mar 2013 11:51:58 +1300, Francois Bissey wrote: 5.7 works fine. :) Thanks Markus </pre> Markus Oehme 2013-03-05T11:15:17 Search the mailing list at Gmane query http://search.gmane.org/?group=$group=gmane.linux.gentoo.science