Mailman 3 March 2006 - SciPy-User

step function
by Jonathan Taylor March 13, 2006

March 13, 2006

i was just wondering if there is a simple way to define a step function in scipy, like an ECDF (empirical cumulative distribution function) in statistics.... i have hacked my own version, but it is quite slow.... thanks, jonathan -- ------------------------------------------------------------------------ I'm part of the Team in Training: please support our efforts for the Leukemia and Lymphoma Society! http://www.active.com/donate/tntsvmb/tntsvmbJTaylor GO TEAM !!! ------------------------------------------------------------------------ Jonathan Taylor Tel: 650.723.9230 Dept. of Statistics Fax: 650.725.8977 Sequoia Hall, 137 www-stat.stanford.edu/~jtaylo 390 Serra Mall Stanford, CA 94305

2 3

GCC 3.3 requirement
by Robert Hetland March 13, 2006

March 13, 2006

Apple shipped a crippled version of GCC 3.3 with the new intel Macs. (It seems like it is there -- there is a package, and you can gcc_select it -- but it does not work properly.) Many on this list have discouraged using 4.0 for building scipy. I have also found that 3.3 works on the *PPC* Mac, where 4.0 does not. However, it seems inevitable that mac users will need to compile with 4.0 as more users get new Intel based machines. I am sure that eventually someone will provide an unofficial gcc 3.3 package (like Guarav Khanna who maintains hpc.sf.net), but I don't think this is a good long-term solution. What are the issues involved? Why do we need gcc 3.3? Is there some sort of timescale for moving the code base up to GCC 4.0 (or 4.x) compliance? -Rob. ----- Rob Hetland, Assistant Professor Dept of Oceanography, Texas A&M University p: 979-458-0096, f: 979-845-6331 e: hetland(a)tamu.edu, w: http://pong.tamu.edu

4 5

ANN: PyDSTool v0.83 - new release.
by Robert Clewley March 12, 2006

March 12, 2006

Dear SciPy users, PyDSTool is an integrated simulation, modeling and analysis package for dynamical systems (including ODEs, DAEs, maps, and hybrid systems) and scientific data. Building on SciPy classes, the package also supports symbolic expression processing, bifurcation analysis, and enhanced arrays for "index-free" and highly contextualized scientific data manipulation. Model building tools are provided, which use symbolic expression and hierarchical specification classes to ease the development and analysis of complex dynamical models. This includes automated compilation of symbolic representations of models into fast numerical code using enhanced legacy Fortran and C integrators for both stiff and non-stiff systems. A full overview and extensive user documentation is provided online at http://pydstool.sourceforge.net, and in the code itself. We have made a significant update in version 0.83 of PyDSTool, including added features for bifurcation/continuation analysis, symbolic expression manipulation and symbolic differentiation, use of Points as enhanced arrays, and the addition of support for external inputs to ODEs, maps, etc. (e.g., from experimental data sources). A full list of what's new is provided in the release notes on SourceForge. Thanks for your attention, Rob Clewley, Erik Sherwood, Drew LaMar, Dept. of Mathematics and Center for Applied Mathematics, Cornell University.

2 2

Re: [SciPy-user] GCC 3.3 requirement
by Webb Sprague March 12, 2006

March 12, 2006

Hi All, Just for the record, I used GCC-3.4.5 on a gentoo box recently to build scipy, and ran into a bug wherein singular value decomposition hangs forever, so I rev'ed back to GCC-3.3.x. I just worked around it, but it might be worth looking into http://bugs.gentoo.org/show_bug.cgi?id=114885 W

1 0

Re: [SciPy-user] GCC 3.3 requirement
by Bill Northcott March 11, 2006

March 11, 2006

On 11/03/2006, at 1:14 AM, Robert Hetland wrote: > Apple shipped a crippled version of GCC 3.3 with the new intel > Macs. (It seems like it is there -- there is a package, and you can > gcc_select it -- but it does not work properly.) Apple are very clear that their gcc-3.3 does NOT support the Intel machines. It is not crippled, it is just PPC software running under Rosetta! That is why everyone is working to get packages to function with gcc-4. Bill Northcott

1 0

error compiling scipy with gcc 4.1
by Guillem Borrell Nogueras March 10, 2006

March 10, 2006

Hi I am trying to compile scipy svn (r1676) with gcc 4.1. One of gfortran's developers suggested me to use gcc 4.1 instead of 4.0.2 because this release is supposed to be much more stable and fast. After building numpy successfully, also from svn, the scipy compilation fails when tries to build the dfftpack library. It seems that distutils fails to find the fortran compiler. I'm a bit perplexed about this because it has been able to build all the previous fortran subpackages... ... customize IntelFCompiler customize LaheyFCompiler customize PGroupFCompiler customize AbsoftFCompiler customize NAGFCompiler customize VastFCompiler customize GnuFCompiler customize CompaqFCompiler customize IntelItaniumFCompiler customize Gnu95FCompiler customize G95FCompiler customize GnuFCompiler customize Gnu95FCompiler customize GnuFCompiler customize GnuFCompiler using build_clib building 'dfftpack' library compiling Fortran sources f77(f77) options: '-Wall -fno-second-underscore -fPIC -O2 -funroll-loops -fomit-frame-pointer -malign-double' compile options: '-c' f77:f77: Lib/fftpack/dfftpack/dcost.f sh: f77: command not found sh: f77: command not found error: Command "f77 -Wall -fno-second-underscore -fPIC -O2 -funroll-loops -fomit-frame-pointer -malign-double -c -c Lib/fftpack/dfftpack/dcost.f -o build/temp.linux-i686-2.4/Lib/fftpack/dfftpack/dcost.o" failed with exit status 127 I am not specially interested in building scipy with gcc 4.1 but I thought that the resoults would have some interest. guillem -- Guillem Borrell Nogueras WEBSITE http://torroja.dmt.upm.es:9673/Guillem_Site/ BLOG http://torroja.dmt.upm.es:9673/Guillem_Borrell/ EMAIL guillemborrell_at_gmail.com (personal) guillem_at_torroja.dmt.upm.es (CFD Lab. ETSIA)

1 0

SciPy job ad
by Gennan Chen March 8, 2006

March 8, 2006

Hi! All, I know this is not the right place to put an ad. But I think someone in scipy community might be interested in this job: Neuroimaging System Engineer/Programmer CorTechs Labs, Inc. is developing MRI-based diagnostics which are targeted at the volumetric measurement, classification, and early detection of specific Central Nervous System diseases, including Alzheimer’s disease, Multiple Sclerosis, and HIV. We are currently seeking candidates for a Neuroimaging System Engineer/Programmer position which will develop and support commercial software code which allows radiolical and healthcare IT components to interface with our diagnostic software components. As a member of a small development team, you will work closely with world-class scientists and physicians who are participating in the National Institute of Health/National Institute of Aging/National Institute of Infectious Diseases sponsored projects. Qualifications -3-5 years of experience in developing commercial and or academic applications in the Python language. -Expert in building Python extension module -Experience in embedding Python -3 – 5 years of experience in DICOM components and a thorough understanding of the DICOM standard -3 – 5 years of experience with Matlab and C/C++ -Experience with HPC (High Performance Computation) system -Experience with LAMP (Linux, Apache, MySQL +Python/PHP) -Experience with SQL programming calls to relational database management systems -Experience with QT/PyQT and OpenGL will be a plus. Duties and Responsibilities -Based upon written specifications and perhaps well defined prototype components, design and build commercial interface code which allows clinicians to use our computationally derived products within workflow customary in hospitals and radiology departments. -Develop a service based extension which allows for broader use of a centrally available computational resource for medical researchers -Help scientist to port current code to Python-based scientific computation environment. Education A M.S. or higher from an accredited college in Computer Science or other related discipline, and 3.5 or higher GPA is required. Cortechs Labs, Inc. is located in La Jolla Village in San Diego, CA. This is a full-time position which includes a competitive salary, subsidized healthcare benefits, vacation, and if needed H1B sponsorship. Qualified candidates should send or email (please do not send Word files- they will not be opened) their resumes to the attention of: Howard Pinsky, Chief Operating Officer Cortechs Labs, Inc. 1020 Prospect St., Suite 304 La Jolla, CA 92037 Or email to: pinsky(a)cortechs.net Gen-Nan Chen, PhD Chief Scientist Research and Development Group CorTechs Labs Inc (www.cortechs.net) 1020 Prospect St., #304, La Jolla, CA, 92037 Tel: 1-858-459-9700 ext 16 Fax: 1-858-459-9705 Email: gnchen(a)cortechs.net

1 0

numpy/scipy+ATLAS on SuSE10.0/gcc4 (once more)
by Christian Kristukat March 8, 2006

March 8, 2006

Hi, I've got some observations to share concerning the compilation of numpy/scipy with ATLAS on SuSE10.0 In some earlier posts I told everybody to use gfortran to build numpy/scipy/ATLAS. Although the compilation worked without errors and numpy/scipy seemed to work as well, Robert prooved, that gfortran is causing serious faults so that scipy eventually fails during execution. Originally I have chosen to use gfortran because every attempt to build numpy/scipy/ATLAS on SuSE10.0 using gcc4 and g77 failed. I will shortly describe what I tried (with gcc4/g77) 1) using SUSE10.0 blas and lapack rpms Python 2.4.1 (#1, Sep 13 2005, 00:39:20) [GCC 4.0.2 20050901 (prerelease) (SUSE Linux)] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> import numpy import linalg -> failed: /usr/lib/libblas.so.3: undefined symbol: _gfortran_filename >>> 2) with ATLAS Python 2.4.1 (#1, Sep 13 2005, 00:39:20) [GCC 4.0.2 20050901 (prerelease) (SUSE Linux)] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> import numpy import linalg -> failed: /usr/lib/python2.4/site-packages/numpy/linalg/lapack_lite.so: undefined symbol: _gfortran_copy_string >>> 3) using self compiled lapack and blas works!! Can anyone figure out what the problem is and propose what to try next? Regards, Christian

3 7

Compiling scipy with ifort on an intel Mac
by Rob Hetland March 8, 2006

March 8, 2006

I have had a hard time getting g77 compiled on my intel Mac, but I do have a beta version of the intel fortran compiler for the Mac... So, I am trying to get scipy compiled with ifort. I get an error like that shown below. I get a similar error when trying to use f2py. I should note that numpy (other than f2py) compiles fine. I am such a hack when it comes to these things -- I'm sure I am doing something stupid. I have tried a number of things -- I have the regular Mac python, and a fresh install in /usr/local, but both fail in the same way. I have tried including/excluding some of the obvious compiler flags such as -nofor_main. It seems that the linker can't find the python libraries. Including -lpython reduces the list of undefined symbols, but _MAIN__ still remains. Can someone point me to the truth? -Rob EXAMPLE: mire:~/src/python/scipy$ python setup.py config_fc --fcompiler=intel build [...snip...] f2py options: [] adding 'build/src/fortranobject.c' to sources. adding 'build/src' to include_dirs. building extension "scipy.special._cephes" sources building extension "scipy.special.specfun" sources f2py options: ['--no-wrap-functions'] adding 'build/src/fortranobject.c' to sources. adding 'build/src' to include_dirs. building extension "scipy.stats.statlib" sources f2py options: ['--no-wrap-functions'] adding 'build/src/fortranobject.c' to sources. adding 'build/src' to include_dirs. building extension "scipy.stats.futil" sources f2py options: [] adding 'build/src/fortranobject.c' to sources. adding 'build/src' to include_dirs. building extension "scipy.stats.mvn" sources f2py options: [] adding 'build/src/fortranobject.c' to sources. adding 'build/src' to include_dirs. adding 'build/src/Lib/stats/mvn-f2pywrappers.f' to sources. building data_files sources running build_py copying Lib/__svn_version__.py -> build/lib.darwin-8.5.2-i386-2.4/scipy copying build/src/scipy/__config__.py -> build/lib.darwin-8.5.2- i386-2.4/scipy running build_clib customize UnixCCompiler customize UnixCCompiler using build_clib Could not locate executable efort Could not locate executable efc customize IntelFCompiler customize IntelFCompiler using build_clib running build_ext customize UnixCCompiler customize UnixCCompiler using build_ext customize IntelFCompiler customize IntelFCompiler using build_ext building 'scipy.interpolate._fitpack' extension compiling C sources gcc options: '-fno-strict-aliasing -Wno-long-double -no-cpp-precomp - mno-fused-madd -DNDEBUG -g -O3 -Wall -Wstrict-prototypes' compile options: '-I/usr/local/lib/python2.4/site-packages/numpy/core/ include -I/usr/local/include/python2.4 -c' /opt/intel/fc/9.1.017/bin/ifort -shared -nofor_main build/ temp.darwin-8.5.2-i386-2.4/Lib/interpolate/_fitpackmodule.o -Lbuild/ temp.darwin-8.5.2-i386-2.4 -lfitpack -o build/lib.darwin-8.5.2- i386-2.4/scipy/interpolate/_fitpack.so ifort: Command line warning: ignoring unknown option '-shared' ld: Undefined symbols: _PyArg_ParseTuple _PyCObject_AsVoidPtr _PyCObject_Type _PyDict_SetItemString _PyErr_Format _PyErr_NewException _PyErr_NoMemory _PyErr_Occurred _PyExc_RuntimeError _PyImport_ImportModule _PyModule_GetDict _PyObject_GetAttrString _PyString_FromString _Py_BuildValue _Py_FatalError _Py_InitModule4 _MAIN__ ifort: Command line warning: ignoring unknown option '-shared' ld: Undefined symbols: _PyArg_ParseTuple _PyCObject_AsVoidPtr _PyCObject_Type _PyDict_SetItemString _PyErr_Format _PyErr_NewException _PyErr_NoMemory _PyErr_Occurred _PyExc_RuntimeError _PyImport_ImportModule _PyModule_GetDict _PyObject_GetAttrString _PyString_FromString _Py_BuildValue _Py_FatalError _Py_InitModule4 _MAIN__ error: Command "/opt/intel/fc/9.1.017/bin/ifort -shared -nofor_main build/temp.darwin-8.5.2-i386-2.4/Lib/interpolate/_fitpackmodule.o - Lbuild/temp.darwin-8.5.2-i386-2.4 -lfitpack -o build/lib.darwin-8.5.2- i386-2.4/scipy/interpolate/_fitpack.so" failed with exit status 1 mire:~/src/python/scipy$ python setup.py config_fc --fcompiler=intel build

3 2

minimizers don't work - d1mach problem
by Olivier Médoc March 7, 2006

March 7, 2006

I found this thread in the mailinglist archive. I seems to get the same error : scipy 0.4.6 build with a gcc4.0.3 snapshot and gfortran4.0.3 snapshot. When I run test() I get : Adjust D1MACH by uncommenting data statements appropriate for your machine. STOP 779 I think this error is an arch specific error (AMD64 ?). Finally I found a workaround. I can run the test suite at level 10 without errors : - I deleted the two d1mach.f files and replaced them with a cfile found here : |https://pse.cheme.cmu.edu/svn-view/ascend/code/trunk/linpack/ (the file d1mach.c) | - then adjusted the setup.py files | sed "s/join('mach','\*\.f')/join('mach','*.f'),join('mach','*.c')/" -i Lib/special/setup.py sed "s/join('mach','\*\.f')/join('mach','*.f'),join('mach','*.c')/" -i Lib/integrate/setup.py |Please note that I considered that it worked because I successfully compute the tests at level 10. But I don't know more about this.. Regards Olivier. ___________________________________________________________________________ Nouveau : téléphonez moins cher avec Yahoo! Messenger ! Découvez les tarifs exceptionnels pour appeler la France et l'international. Téléchargez sur http://fr.messenger.yahoo.com

1 0