Mailman 3 August 2003 - SciPy-User

pointing to Numeric source
by John A. Turner 08 Aug '03

08 Aug '03

I have Numeric installed in a non-standard place, so my SciPy build (0.2.0-200.4161) is failing with: building 'scipy_base.fastumath' extension cc -DNDEBUG -O -Olimit 1500 -I/usr/local/include/python2.2 -c /home/turner/Tools/src/Python/SciPy-0.2.0_alpha_200.4161/scipy_core/scipy_base/fastumathmodule.c -o build/temp.osf1-V5.1-alpha-2.2/fastumathmodule.o cc: Severe: /home/turner/Tools/src/Python/SciPy-0.2.0_alpha_200.4161/scipy_core/scipy_base/fastumathmodule.c, line 2: Cannot find file "Numeric/arrayobject.h" specified in #include directive. (noinclfilef) #include "Numeric/arrayobject.h" -^ error: command 'cc' failed with exit status 1 how can I tell setup.py where to look for the Numeric source? thx, -John A. Turner Los Alamos National Lab Advanced Scientific Simulation, CCS-2

2 2

Sparse matrices?
by Gerber G <ggerber＠sun.ac.za> 08 Aug '03

08 Aug '03

Hi, I am currently programming Finite Elements. Does scipy support sparse matrices? Regards, George

2 1

scipy AssertionError
by Ulrich Guenther 08 Aug '03

08 Aug '03

On a linux installation of scipy I get an error when running scipy.test(): . ====================================================================== FAIL: check_basic (test_morestats.test_shapiro) ---------------------------------------------------------------------- Traceback (most recent call last): File "/usr/lib/python2.2/site-packages/scipy/stats/tests/test_morestats.py", line 37, in check_basic assert_almost_equal(pw,0.52459925413131714,6) File "/usr/lib/python2.2/site-packages/scipy_test/testing.py", line 349, in assert_almost_equal assert round(abs(desired - actual),decimal) == 0, msg AssertionError: Items are not equal: DESIRED: 0.52459925413131714 ACTUAL: 0.52460002899169922 ---------------------------------------------------------------------- Ran 739 tests in 3.160s FAILED (failures=1) <unittest.TextTestRunner instance at 0x87b9364> Where can this come from? My config: suse 8.2, kernel 2.4.20, python 2.2, gcc-3.3, lapack with atlas Ulrich

2 1

build prob (and fix) on OSF
by John A. Turner 08 Aug '03

08 Aug '03

building 0.2.0-200.4161 on an Alpha running Tru64Unix 5.1 (also known as OSF1, and which sys.platform reports as osf1V5), ran into a failure after a huge ar command for LAPACK went hunting and found this bit in build_flib.py: if os.name == 'nt' or sys.platform[:4] == 'irix': # I (pearu) had the same problem on irix646 ... # I think we can make this "bunk" default as skip_ranlib # feature speeds things up. # XXX:Need to check if Digital compiler works here. # This is pure bunk... # Windows fails for long argument strings on the command line. # if objects is real long (> 2048 chars or so on my machine), # the command fails (cmd.exe /e:2048 on w2k). # for now we'll split linking into to steps which should work for objects = object_list[:] while objects: sure enough, changing the if test to: if os.name == 'nt' or sys.platform[:4] == 'irix' or sys.platform[:3] == 'osf': allowed me to get past that problem -John A. Turner Los Alamos National Lab Advanced Scientific Simulation, CCS-2

1 0

scipy AssertionError
by Ulrich Guenther 07 Aug '03

07 Aug '03

On a linux installation of scipy I get an error when running scipy.test(): . ====================================================================== FAIL: check_basic (test_morestats.test_shapiro) ---------------------------------------------------------------------- Traceback (most recent call last): File "/usr/lib/python2.2/site-packages/scipy/stats/tests/test_morestats.py", line 37, in check_basic assert_almost_equal(pw,0.52459925413131714,6) File "/usr/lib/python2.2/site-packages/scipy_test/testing.py", line 349, in assert_almost_equal assert round(abs(desired - actual),decimal) == 0, msg AssertionError: Items are not equal: DESIRED: 0.52459925413131714 ACTUAL: 0.52460002899169922 ---------------------------------------------------------------------- Ran 739 tests in 3.160s FAILED (failures=1) <unittest.TextTestRunner instance at 0x87b9364> Where can this come from? My config: suse 8.2, kernel 2.4.20, python 2.2, gcc-3.3, lapack with atlas Ulrich

1 0

scipy AssertionError
by Ulrich Guenther 07 Aug '03

07 Aug '03

On a linux installation of scipy I get an error when running scipy.test(): . ====================================================================== FAIL: check_basic (test_morestats.test_shapiro) ---------------------------------------------------------------------- Traceback (most recent call last): File "/usr/lib/python2.2/site-packages/scipy/stats/tests/test_morestats.py", line 37, in check_basic assert_almost_equal(pw,0.52459925413131714,6) File "/usr/lib/python2.2/site-packages/scipy_test/testing.py", line 349, in assert_almost_equal assert round(abs(desired - actual),decimal) == 0, msg AssertionError: Items are not equal: DESIRED: 0.52459925413131714 ACTUAL: 0.52460002899169922 ---------------------------------------------------------------------- Ran 739 tests in 3.160s FAILED (failures=1) <unittest.TextTestRunner instance at 0x87b9364> Where can this come from? My config: suse 8.2, kernel 2.4.20, python 2.2, gcc-3.3, lapack with atlas Ulrich

1 0

Newbie type questions
by Jeffrey B. Layton 07 Aug '03

07 Aug '03

Hello, I've played with Scipy off and on for a few months, but now I need to get serious. I am co-authoring an engineering book and will have a large number of examples in it. We are already using MATLAB (and Octave), but I have convinced my co-authors and perhaps the publishers to put Python/Scipy in there as well. This text is a graduate level text and the models in the book can be quite large and time-consuming. Towards this, we have an appendix where we explain the basics of MATLAB and Octave for those who need to brush up or who need a quick intro (enough to run the examples). I'm writing the companion appendix for Python/Scipy. So, I'm working through the examples. This leads me to my questions If I declare a complex matrix A, A=array([[complex(1,1),complex(2,2)],[complex(3,3),complex(4,4)]]) and I want to take the transpose, which I'm assuming to be the complex conjugate, I need to type AT=transpose(conjugate(A)) to get the actual complex conjugate. Is there an easier way to do this? A somewhat related question, if I declare a vector X, X=array([1,2,3]) and then try to take the transpose of it, Y=transpose(X) then X=Y. I sort of understand this since the shape of X is (3, ). Shouldn't the shape of X be (3,1)? (I'm sorry this is such a newbie question). Warning: I'm still working my way through the appendix and I'm re-reading the tutorials on the scipy website, so assume I'll be asking more questions =:) (I'll try not to be a pest). Thanks very much! Jeff

3 4

numarray.dot
by Colin J. Williams 06 Aug '03

06 Aug '03

This function appears to take only partial account of imaginary data, see the second exeution below.. I would have expected a complex element type in a ComplexArray instance. or a message that this type of input cannot be handled. Colin W. print 'version:', _sys.version import numarray as N z= N.dot([1, 2], [[3], [4]]) print 'z, type(z), z[0], z.type():', z, type(z), z[0], z.type() z= N.dot([1, 2], [[3j], [4]]) print 'z, type(z), z[0], z.type():', z, type(z), z[0], z.type() Results: version: 2.3c2 (#45, Jul 24 2003, 21:23:54) [MSC v.1200 32 bit (Intel)] z, type(z), z[0], z.type(): [11] <class 'numarray.numarraycore.NumArray'> 11 Int32 z, type(z), z[0], z.type(): [ 8.] <class 'numarray.numarraycore.NumArray'> 8.0 Float64

2 6

scipy.stats on OSX
by Christopher Fonnesbeck 03 Aug '03

03 Aug '03

There seems to be a platform-specific problem with the stats module within scipy. Specifically, many distributions do not work on OSX builds within distributions.py. For example, any value passed to gamma.pdf returns a NaN, rather than the expected likelihood value. I am using the latest cvs. cjf -- Christopher J. Fonnesbeck (chris at fonnesbeck dot org) GA Coop. Fish & Wildlife Research Unit, University of Georgia

1 0

ranlib?
by Heberto del Rio 02 Aug '03

02 Aug '03

Is there any way to replace the ranlib module (since it is only used as a pseudo random number generator) and replace it with a much better one: Mersenne Twister pseudo random number generator? Heberto +--------------------------------------------------------------------------+ | | | I am in the process of changing ISP, therefore my email address will | | | | change too. Please use the following email address in the future: | | | | hebertodelrio(a)fastmail.fm | | | +--------------------------------------------------------------------------+

1 0