Re: [freeeos-general] eos test
Brought to you by:
airwin
From: Alan W. I. <ir...@be...> - 2008-11-28 21:00:24
|
On 2008-11-25 14:36-0800 Alan W. Irwin wrote: [...] > To get access to my present code, please download the svn trunk version > of FreeEOS. To do that please run the following command: > > svn checkout https://freeeos.svn.sourceforge.net/svnroot/freeeos/trunk \ > freeeos_svn > > where freeeos_svn is the (arbitrary) name of the directory where you want > to store the svn trunk version of my source code. > > I then recommend (see README) you do the build using an initially empty > build tree. > > # Dangerous command if you misname the directory! > rm -rf build_dir_svn > mkdir build_dir_svn > cd build_dir_svn > > # Note ../freeeos_svn is the clean source tree you generated with the > # svn checkout command above. Also, "whatever_prefix" is the (arbitrary) > # name of the install prefix you will use for the "make install" command > # below. > cmake -DCMAKE_INSTALL_PREFIX=whatever_prefix ../freeeos_svn >& cmake.out > make VERBOSE=1 >& make.out > > # Will install in "whatever_prefix" directory so be careful of what you specify > # above. > make VERBOSE=1 install >& make_install.out > > Please let me know if you have any troubles with the build/install. > > Furthermore, once that goes well let me know if the updated code solves all > convergence troubles you encountered for the kif=-1 case. Once I get good > test results from you (and anybody else who wants to help me out by testing > kif=-1 or kif=1), I will make an official FreeEOS-2.2.2 release. Aaron reports (see my previous post) good results for the svn trunk version, kif=1, and a wide variety of free-energy models. Thus, as soon as Santi is able to give me a good report as well for kif=-1, that should resolve all known FreeEOS convergence robustness issues. This weekend, I plan to implement a much more extensive series of tests for FreeEOS that each user can optionally run to make sure that FreeEOS is working (and converging) properly for a wide variety of conditions for their platform. This test series should catch any future convergence problems similar to those found by Aaron and Santi, and therefore should help to make further releases of FreeEOS (I hope to release 2.2.2 next week) much more reliable than previous releases. Alan __________________________ Alan W. Irwin Astronomical research affiliation with Department of Physics and Astronomy, University of Victoria (astrowww.phys.uvic.ca). Programming affiliations with the FreeEOS equation-of-state implementation for stellar interiors (freeeos.sf.net); PLplot scientific plotting software package (plplot.org); the libLASi project (unifont.org/lasi); the Loads of Linux Links project (loll.sf.net); and the Linux Brochure Project (lbproject.sf.net). __________________________ Linux-powered Science __________________________ |