Mailman 3 March 2010 - NumPy-Discussion

[OT] Starving CPUs article featured in IEEE's ComputingNow portal
by Francesc Alted 22 Mar '10

22 Mar '10

Hi, Konrad Hinsen has just told me that my article "Why Modern CPUs Are Starving and What Can Be Done About It", which has just released on the March/April issue of "Computing in Science and Engineering", also made into this month's free-access selection on IEEE's ComputingNow portal: http://www.computer.org/portal/web/computingnow http://www.computer.org/portal/web/computingnow/0310/whatsnew/cise On it, I discuss one of my favourite subjects, memory access, and why it is important for nowadays computing. There are also recommendations for people wanting to squeeze the maximum of performance out of their computers. And, although I tried to be as language-agnostic as I could, there can be seen some Python references here and there :-). Well, sorry about this semi-OT but I could not resist :-) -- Francesc Alted

9 16

Help!!! Docstrings overrun by markup crap.
by Charles R Harris 22 Mar '10

22 Mar '10

Example, Compute the qr factorization of a matrix. > > Factor the matrix `a` as `qr`, where `q` is orthonormal > (:math:`dot( q_{:,i}, q_{:,j}) = \delta_{ij}`, the Kronecker delta) and > `r` is upper-triangular. > Arrggghhhh... Totally. Unreadable. Why not say the columns of q are orthonormal vectors and r is upper triangular? The math tutorial should go in the notes, if anywhere. Might mention that this is a 'thin' factorization (Golub). Let me propose a rule: no math markup in the summary, ever. Parameters > ---------- > a : array_like, shape (M, N) > Matrix to be factored. > mode : {'full', 'r', 'economic'} > Specifies the information to be returned. 'full' is the default. > mode='r' returns a "true" `r`, while 'economic' returns a > "polluted" > `r` (albeit slightly faster; see Returns below). > > Oh, come now, "true", "polluted"? Sounds a bit political... Actually, 'economic' contains info on the Householder reflections. In any case, why mention it at all, just refer to the return documentation. And wouldn't values be a better word than information? > Returns > ------- > * If mode = 'full': > > * q : ndarray of float or complex, shape (M, K) > * r : ndarray of float or complex, shape (K, N) > > Size K = min(M, N) > > * If mode = 'r': > > * r : ndarray of float or complex, shape (K, N) > > * If mode = 'economic': > > * a2 : ndarray of float or complex, shape (M, N) > > The diagonal and the upper triangle of a2 contains r, > while the rest of the matrix is undefined. > WTF? I'm seeing stars. I may be old and crotchety, and I don't mean to be mean to the folks who have done the hard work to bring the docstring to its current state, but I think things have gotten out of hand. Chuck

6 33

using sunperf for compiling scipy and numpy
by Wolfgang Kerzendorf 22 Mar '10

22 Mar '10

Dear all, I would like to use the sunperf libraries when compiling scipy and numpy. I tried using setupscons.py which seems to check from SUNPERF libraries, but it didnt recognize where mine are: here is a listing of /pkg/linux/SS12/sunstudio12.1 (thats where the sunperf library lives): wkerzend@mosura:/home/wkerzend>ls /pkg/linux/SS12/sunstudio12.1/lib/ CCios/ libdbx_agent.so@ libsunperf.so.3@ amd64/ libfcollector.so@ libtha.so@ collector.jar@ libfsu.so@ libtha.so.1@ dbxrc@ libfsu.so.1@ locale/ debugging.so@ libfui.so@ make.rules@ er.rc@ libfui.so.1@ rw7/ libblacs_openmpi.so@ librtc.so@ sse2/ libblacs_openmpi.so.1@ libscalapack.so@ stlport4/ libcollectorAPI.so@ libscalapack.so.1@ svr4.make.rules@ libcollectorAPI.so.1@ libsunperf.so@ tools_svc_mgr@ ------------------------------- I tried to specify this directory in sites.cfg, but I still get the following errors: Checking if g77 needs dummy main - MAIN__. Checking g77 name mangling - '_', '', lower-case. Checking g77 C compatibility runtime ...-L/usr/lib/gcc/x86_64-redhat-linux/3.4.6 -L/usr/lib/gcc/x86_64-redhat-linux/3.4.6 -L/usr/lib/gcc/x86_64-redhat-linux/3.4.6/../../../../lib64 -L/usr/lib/gcc/x86_64-redhat-linux/3.4.6/../../.. -L/lib/../lib64 -L/usr/lib/../lib64 -lfrtbegin -lg2c -lm Checking MKL ... Failed (could not check header(s) : check config.log in build/scons/scipy/integrate for more details) Checking ATLAS ... Failed (could not check header(s) : check config.log in build/scons/scipy/integrate for more details) Checking SUNPERF ... Failed (could not check symbol cblas_sgemm : check config.log in build/scons/scipy/integrate for more details)) Checking Generic BLAS ... yes Checking for BLAS (Generic BLAS) ... Failed: BLAS (Generic BLAS) test could not be linked and run Exception: Could not find F77 BLAS, needed for integrate package: File "/priv/manana1/wkerzend/install_dir/scipy-0.7.1/scipy/integrate/SConstruct", line 2: GetInitEnvironment(ARGUMENTS).DistutilsSConscript('SConscript') File "/home/wkerzend/python_coala/numscons-0.10.1-py2.6.egg/numscons/core/numpyenv.py", line 108: build_dir = '$build_dir', src_dir = '$src_dir') File "/priv/manana1/wkerzend/python_coala/numscons-0.10.1-py2.6.egg/numscons/scons-local/scons-local-1.2.0/SCons/Script/SConscript.py", line 549: return apply(_SConscript, [self.fs,] + files, subst_kw) File "/priv/manana1/wkerzend/python_coala/numscons-0.10.1-py2.6.egg/numscons/scons-local/scons-local-1.2.0/SCons/Script/SConscript.py", line 259: exec _file_ in call_stack[-1].globals File "/priv/manana1/wkerzend/install_dir/scipy-0.7.1/build/scons/scipy/integrate/SConscript", line 15: raise Exception("Could not find F77 BLAS, needed for integrate package") error: Error while executing scons command. See above for more information. If you think it is a problem in numscons, you can also try executing the scons command with --log-level option for more detailed output of what numscons is doing, for example --log-level=0; the lowest the level is, the more detailed the output it.----- -------------------------- any help is appreciated Wolfgang

1 0

PEP 384
by Neal Becker 22 Mar '10

22 Mar '10

I'm thinking PEP 384, if approved, would require a lot of redesign to numpy.

1 0

Test if one element of string array is in a defined list
by Eric Emsellem 22 Mar '10

22 Mar '10

Hi I would like to test whether strings in a numpy S array are in a given list but I don't manage to do so. Any hint is welcome. ======================================================= # So here is an example of what I would like to do # I have a String numpy array: import numpy as num Sarray = num.asarray(["test1","test2","tutu","toto"]) Farray = num.arange(len(Sarray)) mylist = ["tutu","hello","why"] # and I would like to do: result = num.where(Sarray in mylist, Farray, 0) ======================================================= ===> but of course I get a ValueError since the first test syntax is wrong. I would like to be able to do: Sarray in mylist and get the output: array([False, False, True, False],dtype=bool) since only the 3rd string "tutu" of Sarray is in mylist. Any input is welcome. cheers Eric

2 1

Author NumPy books- Packt Publishing
by Kshipra Singh 22 Mar '10

22 Mar '10

Hi All, I am writing to you for Packt Publishing, the publishers of computer related books. We are planning to extend our catalogue of books based on Scientific Computing Tools and are currently inviting authors interested in writing for Packt. This doesn't need any previous writing experience. Just an expert knowledge of your subject and a passion to share it with others is all that we require. So, if you love NumPy and are interested in authoring a book, please write to us with your book ideas at author(a)packtpub.com. Even if you don't have a book idea and are simply interested in authoring a book, we are still keen to hear from you. More details about the opportunity are available at: http://authors.packtpub.com/content/scientific-compting-tools-write-Packt Thanks Kshipra Singh Author Relationship Manager Packt Publishing www.PacktPub.com Skype: kshiprasingh15 Twitter: http://twitter.com/kshipras Interested in becoming an author? Visit http://authors.packtpub.com for all the information you need about writing for Packt.

1 0

Array bug with character (regression from 1.4.0)
by Ryan May 21 Mar '10

21 Mar '10

The following code, which works with numpy 1.4.0, results in an error: In [1]: import numpy as np In [2]: v = 'm' In [3]: dt = np.dtype('>c') In [4]: a = np.asarray(v, dt) On 1.4.0: In [5]: a Out[5]: array('m', dtype='|S1') In [6]: np.__version__ Out[6]: '1.4.0' On SVN trunk: /home/rmay/.local/lib/python2.6/site-packages/numpy/core/numeric.pyc in asarray(a, dtype, order) 282 283 """ --> 284 return array(a, dtype, copy=False, order=order) 285 286 def asanyarray(a, dtype=None, order=None): ValueError: assignment to 0-d array In [5]: np.__version__ Out[5]: '2.0.0.dev8297' Thoughts? (Filed at: http://projects.scipy.org/numpy/ticket/1436) Ryan -- Ryan May Graduate Research Assistant School of Meteorology University of Oklahoma

3 5

Warnings in numpy.ma.test()
by Pierre GM 19 Mar '10

19 Mar '10

All, As you're probably aware, the current test suite for numpy.ma raises some nagging warnings such as "invalid value in ...". These warnings are only issued when a standard numpy ufunc (eg., np.sqrt) is called on a MaskedArray, instead of its numpy.ma (eg., np.ma.sqrt) equivalent. The reason is that the masked versions of the ufuncs temporarily set the numpy error status to 'ignore' before the operation takes place, and reset the status to its original value. I thought I could use the new __array_prepare__ method to intercept the call of a standard ufunc. After actual testing, that can't work. __array_prepare only help to prepare the *output* of the operation, not to change the input on the fly, just for this operation. Actually, you can modify the input in place, but it's usually not what you want. Then, I tried to use __array_prepare__ to store the current error status in the input, force it to ignore divide/invalid errors and send the input to the ufunc. Doesn't work either: np.seterr in __array_prepare__ does change the error status, but as far as I understand, the ufunc is called is still called with the original error status. That means that if something goes wrong, your error status can stay stuck. Not a good idea either. I'm running out of ideas at this point. For the test suite, I'd suggest to disable the warnings in test_fix_invalid and test_basic_arithmetic. An additional issue is that if one of the error status is set to 'raise', the numpy ufunc will raise the exception (as expected), while its numpy.ma version will not. I'll put also a warning in the docs to that effect. Please send me your comments before I commit any changes. Cheers, P.

9 35

bug in ndarray.resize?
by Alan G Isaac 19 Mar '10

19 Mar '10

Is the zero-fill intentional? If so, it is documented? (NumPy 1.3) Alan Isaac >>> a = np.arange(5) >>> b = a.copy() >>> c = np.resize(a, (5,2)) >>> b.resize((5,2)) >>> c # as expected array([[0, 1], [2, 3], [4, 0], [1, 2], [3, 4]]) >>> b # surprise! array([[0, 1], [2, 3], [4, 0], [0, 0], [0, 0]])

6 7

lists of zeros and ones
by gerardob 19 Mar '10

19 Mar '10

Hello, i would like to produce lists of lists 1's and 0's. For example, to produce the list composed of: L = [[1,0,0,0],[0,1,0,0],[0,0,1,0],[0,0,0,1]] I just need to do the following: n=4 numpy.eye(n,dtype=int).tolist() I would like to know a simple way to generate a list containing all the lists having two 1's at each element. Example, n = 4 L2 = [[1,1,0,0],[1,0,1,0],[1,0,0,1],[0,1,1,0],[0,1,0,1],[0,0,1,1]] Any ideas? Thanks. -- View this message in context: http://old.nabble.com/lists-of-zeros-and-ones-tp27950978p27950978.html Sent from the Numpy-discussion mailing list archive at Nabble.com.

6 7