Mailman 3 September 2007 - SciPy-User

OS X users - Please try multiple scipy.test() runs
by Tom Loredo Oct. 2, 2007

Oct. 2, 2007

Hi folks- I just reported some strange scipy crashes I've experienced on my MacBook running OS X Tiger & Python 2.5.1. I'd like to find out more about the conditions producing the crash. If OS X users could do the following simple test and report the results, it would be very helpful. Please also report details about your platform (machine, OS version, Python, numpy & scipy versions). All you have to do is start Python, import scipy, and run scipy.test() *multiple times*: >>> import scipy >>> scipy.test() [snip] >>> scipy.test() and so on, up to 10 times. For me, with Python-2.5.1, numpy-1.0.3.1, and scipy-0.6.0, right after install I get a crash at the 10th time. Subsequent attempts (with new Python invocations) give crashes earlier and earlier. Eventually the crash is fully reproducible at the 3rd scipy.test(), always crashing here: Found 42 tests for scipy.lib.lapack Found 41 tests for scipy.linalg.basic <module 'scipy.linalg.fblas' from '/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/linalg/fblas.so'> Segmentation fault I get similar crashes with scipy-0.5.2.1, but the conditions are not reproducible. On a G4 (PPC) machine running Py-2.4.4 and scipy-0.5.2.1, I don't get any crashes (I ran scipy.test() 12 times in a row). Your help would be appreciated in sorting out the conditions under which such crashes happen. Thanks, Tom PS: For the curious, I was led to this somewhat accidentally, due to problems with linalg I'm experiencing under RHEL 5 and another user is experiencing with FC4. I was trying to see if the problems were platform-specific. ------------------------------------------------- This mail sent through IMP: http://horde.org/imp/

7 7

Pb with numpy.histogram
by LB Oct. 1, 2007

Oct. 1, 2007

Hi, I've got strange results with numpy.histogram : Here is its doc strings : """ Help on function histogram in module numpy.lib.function_base: histogram(a, bins=10, range=None, normed=False) Compute the histogram from a set of data. :Parameters: - `a` : array The data to histogram. n-D arrays will be flattened. - `bins` : int or sequence of floats, optional If an int, then the number of equal-width bins in the given range. Otherwise, a sequence of the lower bound of each bin. - `range` : (float, float), optional The lower and upper range of the bins. If not provided, then (a.min(), a.max()) is used. Values outside of this range are allocated to the closest bin. - `normed` : bool, optional If False, the result array will contain the number of samples in each bin. If True, the result array is the value of the probability *density* function at the bin normalized such that the *integral* over the range is 1. Note that the sum of all of the histogram values will not usually be 1; it is not a probability *mass* function. :Returns: - `hist` : array (n,) The values of the histogram. See `normed` for a description of the possible semantics. - `lower_edges` : float array (n,) The lower edges of each bin. """ and here is a snipplet of code : >>> r = random.normal(8, 2, 500) >>> r.min(), r.max() (1.164117097856284, 13.069426390055149) >>> ra (3, 12) >>> pdf, xpdf = histogram(r, nbins, range=ra, normed=False) >>> pdf array([ 1, 6, 5, 8, 30, 39, 53, 55, 61, 50, 45, 42, 32, 26, 17, 27]) >>> pdf.sum() 497 It seems I've lost 3 of my 500 random numbers ! >>> r[ r>= ra[1]] array([ 12.00676288, 12.8381615 , 12.48380931, 12.55392835, 12.26153469, 12.92869504, 12.58290343, 12.03782311, 13.06942639, 12.06375346, 12.02970414, 12.53556779, 12.54203654, 12.02611864, 12.85113934, 12.64692817]) >>> r[ r<= ra[0]] array([ 1.1641171 , 2.85873306, 2.92046745]) So this number match the number of experiments below the range given to histogram. This smells like a bug to me. Is there something I've misunderstood in the utilisation of numpy.histogram ? For information >>> numpy.__version__ '1.0.2' Regards, -- LB

5 9

All the roots of a function in an interval
by Gael Varoquaux Oct. 1, 2007

Oct. 1, 2007

Hi, Is there a tool in scipy that will give me all the roots of a function in an interval? I can always make one using optimize.minpack.fslove and a first scan to define different starting points, but if there is one, it would be sweet. Cheers, Gaël

3 4

ANN: Python Equations 2.0 - weaveless version
by zunzun＠zunzun.com Sept. 28, 2007

Sept. 28, 2007

Python Equations 2.0 is available for download at http://sf.net/project/showfiles.php?group_id=187105 This version does not require Weave and only requires python, numpy and scipy. BSD license. About Python Equations The middleware for http://zunzun.com as a collection of Python equations that can fit themselves to both 2D and 3D data sets (curve fitting), output source code in several computing languages, and run a genetic algorithm for initial parameter estimation. James Phillips

4 6

Signification of parameters
by Matthieu Brucher Sept. 28, 2007

Sept. 28, 2007

Hi, Does someone know the meaning of the parameters in stats. f_value_wilks_lambda(ER, EF, dfnum, dfden, a, b), especially a and b ? Matthieu

2 6

scipy.stats.lognorm
by Manu Hack Sept. 28, 2007

Sept. 28, 2007

Hi all, I have a quick question on scipy.stats.lognorm. From the manual it said: Lognormal distribution lognorm.pdf(x,s) = 1/(s*x*sqrt(2*pi)) * exp(-1/2*(log(x)/s)**2) for x > 0, s > 0. At the same time there are loc and scale to control the shape of the distribution. So it's like there are 3 parameters to control the shape of the distribution (but 2 should be enough to specify one). In fact, if we look at R, the function dlnorm, plnorm, etc. are simple and just did the job. And after searching and trying different parameters and compare with R's function, I still couldn't figure out how to use scipy.stats.lognorm. So am I missing anything there? Manu

3 2

Multi-peak fitting
by Clowers, Brian H Sept. 27, 2007

Sept. 27, 2007

After perusing the archives I'm still a bit stumped. I'm looking to create a script in python that will fit multiple peaks to an array of spectroscopic data. It appears as though there are some C++ (fityk) and python (peak-o-mat) programs designed to do this, however, I'd like to incorporate such a function into an automated batch type processing. Ideally, it would nice to access those functions directly but my level of proficiency with python is not quite there yet. Does anyone have an example of how this might be done using the libraries currently available or be willing to share a custom script using the built in functionality of scipy and numpy? My data can be quite simple (i.e. having just one Gaussian peak) or other cases in which many individual and overlapping Gaussian shapes may exist. My current solution is to export the array to a text file and perform the multi-peak fitting using either IGOR Pro or MATLAB. It is my understanding that in order to fit multiple peaks a script designed to identify peaks and provide the initial estimates for the peak fitting is required. I'm not sure how to include such a script with a fitting routine that takes into account all of the potential features in a spectrum. Any help is greatly appreciated. Cheers, Brian

5 5

array TypeError: expected a readable buffer object
by BORGULYA Gábor Sept. 27, 2007

Sept. 27, 2007

Hi List! Could someone explain me why I get an error message for the first expression and why no error for the second? In [1]: import scipy In [2]: a = scipy.array(['1','2','3',None,'4','5']) --------------------------------------------------------------------------- exceptions.TypeError Traceback (most recent call last) /home/gab/<console> TypeError: expected a readable buffer object In [3]: a = scipy.array([1,2,3,None,4,5]) In [4]: a Out[4]: array([1, 2, 3, None, 4, 5], dtype=object) I expected that the first expression would evaluate as array(['1', '2', '3', None, '4', '5'], dtype=object) but it did not. Please Cc my email address when replying to the list, thank you! Gábor

2 3

4 test failures in test_odr for scipy svn 3371
by George Nurser Sept. 27, 2007

Sept. 27, 2007

I compiled latest svn version, using gcc 3.4.4 on RH Linux Opteron 64-bit, and intel fortran v9.1: python setup.py config --fcompiler=intelem build_clib --fcompiler=intelem build_ext --fcompiler=intelem install >& inst.log & I had modified numpy/distutils/intel.py so as to include -xW flag for ifort, as recomended by Intel for the Opteron. scipy.test(level=1) gave four failures, in: test_explicit (scipy.tests.test_odr.test_odr, test_lorentz (scipy.tests.test_odr.test_odr) test_multi (scipy.tests.test_odr.test_odr) test_pearson (scipy.tests.test_odr.test_odr) --George Nurser. ====================================================================== FAIL: test_explicit (scipy.tests.test_odr.test_odr) ---------------------------------------------------------------------- Traceback (most recent call last): File "/noc/users/agn/ext/AMD64/lib/python2.5/site-packages/scipy/odr/tests/test_odr.py", line 50, in test_explicit -8.7849712165253724e-02]), File "/noc/users/agn/ext/AMD64/lib/python2.5/site-packages/numpy/testing/utils.py", line 232, in assert_array_almost_equal header='Arrays are not almost equal') File "/noc/users/agn/ext/AMD64/lib/python2.5/site-packages/numpy/testing/utils.py", line 217, in assert_array_compare assert cond, msg AssertionError: Arrays are not almost equal (mismatch 100.0%) x: array([ 1.26462971e+03, -5.42545890e+01, -8.64250389e-02]) y: array([ 1.26465481e+03, -5.40184100e+01, -8.78497122e-02]) ====================================================================== FAIL: test_lorentz (scipy.tests.test_odr.test_odr) ---------------------------------------------------------------------- Traceback (most recent call last): File "/noc/users/agn/ext/AMD64/lib/python2.5/site-packages/scipy/odr/tests/test_odr.py", line 295, in test_lorentz 3.7798193600109009e+00]), File "/noc/users/agn/ext/AMD64/lib/python2.5/site-packages/numpy/testing/utils.py", line 232, in assert_array_almost_equal header='Arrays are not almost equal') File "/noc/users/agn/ext/AMD64/lib/python2.5/site-packages/numpy/testing/utils.py", line 217, in assert_array_compare assert cond, msg AssertionError: Arrays are not almost equal (mismatch 100.0%) x: array([ 1.00000000e+03, 1.00000000e-01, 3.80000000e+00]) y: array([ 1.43067808e+03, 1.33905090e-01, 3.77981936e+00]) ====================================================================== FAIL: test_multi (scipy.tests.test_odr.test_odr) ---------------------------------------------------------------------- Traceback (most recent call last): File "/noc/users/agn/ext/AMD64/lib/python2.5/site-packages/scipy/odr/tests/test_odr.py", line 191, in test_multi 0.5101147161764654, 0.5173902330489161]), File "/noc/users/agn/ext/AMD64/lib/python2.5/site-packages/numpy/testing/utils.py", line 232, in assert_array_almost_equal header='Arrays are not almost equal') File "/noc/users/agn/ext/AMD64/lib/python2.5/site-packages/numpy/testing/utils.py", line 217, in assert_array_compare assert cond, msg AssertionError: Arrays are not almost equal (mismatch 100.0%) x: array([ 4. , 2. , 7. , 0.4, 0.5]) y: array([ 4.37998803, 2.43330576, 8.00288459, 0.51011472, 0.51739023]) ====================================================================== FAIL: test_pearson (scipy.tests.test_odr.test_odr) ---------------------------------------------------------------------- Traceback (most recent call last): File "/noc/users/agn/ext/AMD64/lib/python2.5/site-packages/scipy/odr/tests/test_odr.py", line 238, in test_pearson np.array([ 5.4767400299231674, -0.4796082367610305]), File "/noc/users/agn/ext/AMD64/lib/python2.5/site-packages/numpy/testing/utils.py", line 232, in assert_array_almost_equal header='Arrays are not almost equal') File "/noc/users/agn/ext/AMD64/lib/python2.5/site-packages/numpy/testing/utils.py", line 217, in assert_array_compare assert cond, msg AssertionError: Arrays are not almost equal (mismatch 100.0%) x: array([ 1., 1.]) y: array([ 5.47674003, -0.47960824]) ---------------------------------------------------------------------- Ran 1771 tests in 4.373s FAILED (failures=4) <unittest._TextTestResult run=1771 errors=0 failures=4>

1 0

EMpy
by lorenzo bolla Sept. 27, 2007

Sept. 27, 2007

I'm very pleased to announce the release of EMpy (Electromagnetic Python), a suite of numerical algorithms widely used in electromagnetism. The package, in its very-alpha stage by now, only contains the transfer matrix and the rigorous coupled wave analysis algorithms, and some handy functions and classes (to model materials, for examples). The idea is to expand it in the near future with new algorithms and interfaces to existent software. Package's homepage: http://empy.sourceforge.net Anyone interested in contributing, please contact me: lbolla(a)users.sourceforge.net. The package is based on Numpy/Scipy.

1 0