Mailman 3 November 2011 - NumPy-Discussion

magma?
by Neal Becker 18 Nov '11

18 Nov '11

Is there any possibility of incorporating this work into numpy? http://icl.cs.utk.edu/magma/software/index.html

3 2

BSD C port of FFTPACK incl. bluestein algorithm
by Dag Sverre Seljebotn 18 Nov '11

18 Nov '11

I've been in touch with Martin Reinecke, author of the libpsht code for spherical harmonic transforms, about licensing issues. libpsht itself will remain under the GPL, but he is likely to release his C port of FFTPACK under BSD in the near future, as it is based on the public domain FFTPACK. I'm grateful for this change for my own purposes (allows releasing my own competing SHT library under the BSD) -- but it could perhaps be useful for NumPy or SciPy as well, depending on how complete the port is? E.g., perhaps make numpy.fft more complete (is the numpy.fft/scipy.fftpack split simply because of the Fortran dependency?). Not sure about whether all of FFTPACK or a subset is included, but it does include a Bluestein implementation in addition. http://libpsht.svn.sourceforge.net/viewvc/libpsht/libfftpack/ Dag Sverre

4 6

Re: [Numpy-discussion] how to use the name of a ndarray as a string
by Chao YUE 18 Nov '11

18 Nov '11

Thanks to all people for this very nice discussions. the solutions are more that what I want!! and help me to clarify some concepts, and really begin to use class as a beginner :) I would say either Olivier or denis's solution can solve my problem completely. cheers, Chao 2011/11/18 denis <denis-bz-gg(a)t-online.de> > as Olivier and Chris say, use a dict from the start. > To make this a bit easier / more readable, > use this tiny class Dotdict: > > class Dotdict( dict ): > """ d = dotdict(), d.key same as d["key"] """ > # aka Bag, Bunch > # cPickle.dumps( d, -1 ): PicklingError in 2.6.4 > def __init__(self, **kwargs): > dict.__init__(self, kwargs) > self.__dict__ = self > > d = Dotdict() > d.a = np.arange(10) > d.b = np.ones((3,4)) > d.b[1,2] += 1 > print d.b > > > One can also save all the d.* arrays like this: > > def savedotdict( d, prefix="d.", save=np.savetxt, **kwargs ): > """ save all d.* to files """ > for name, val in d.items(): > out = prefix + name > if isinstance( val, basestring ): > with open( out, "w" ) as f: > f.write( val.rstrip() ) > f.write( "\n" ) > else: > print "saving %s to %s" % (val.shape, out) > save( out, val, **kwargs ) > > d.info = "2011-11-18 nov kilroy" # From() > savedotdict( d, "d.", fmt="%.3g" ) > > cheers > -- denis > > (If you use this, could you post it to the numpy-discussion group > please ? > It rejects me, grr.) > > > On Nov 10, 11:17 am, Chao YUE <chaoyue...(a)gmail.com> wrote: > > Hi, > > > > Does anyone know how I can quickly use the name of a ndarray as a string? > -- *********************************************************************************** Chao YUE Laboratoire des Sciences du Climat et de l'Environnement (LSCE-IPSL) UMR 1572 CEA-CNRS-UVSQ Batiment 712 - Pe 119 91191 GIF Sur YVETTE Cedex Tel: (33) 01 69 08 29 02; Fax:01.69.08.77.16 ************************************************************************************

2 1

Subclassing ndarray
by Fabrice Silva 18 Nov '11

18 Nov '11

Hi, I wish I could subclass ndarray so that I can use the ufuncs, but that I could ensure the array remains Fortran contiguous, and that the number of dimension is always 3 (even if the dtype can be more complex than a simple numerical type). My intention is to deal with fields that can be scalar, vectorial or even tensorial, but it must stay compatible with a library we are using that assumes 3D problems with column-major order, and some more attributes concerning spatial steps, strides and comments... I have been looking at __array_{wrap, prepare, finalize}__ methods. I would believe __array_prepare__ would be the good starting point to impose the 'F'-ordered array and atleast_3d, but the point is that I don't know at that time if the ufunc will transform the datatype (from vector to scalar, from int to float, etc...) Some thought? Regards -- Fabrice Silva

1 0

ANN: Spyder v2.1.2
by Pierre Raybaut 16 Nov '11

16 Nov '11

Hi all, On the behalf of Spyder's development team (http://code.google.com/p/spyderlib/people/list), I'm pleased to announce that Spyder v2.1.2 has been released and is available for Windows XP/Vista/7, GNU/Linux and MacOS X: http://code.google.com/p/spyderlib/ As this is mostly a maintenance release, a lot of bugs were fixed and some minor features were added: http://code.google.com/p/spyderlib/wiki/ChangeLog Spyder is a free, open-source (MIT license) interactive development environment for the Python language with advanced editing, interactive testing, debugging and introspection features. Originally designed to provide MATLAB-like features (integrated help, interactive console, variable explorer with GUI-based editors for dictionaries, NumPy arrays, ...), it is strongly oriented towards scientific computing and software development. Thanks to the `spyderlib` library, Spyder also provides powerful ready-to-use widgets: embedded Python console (example: http://packages.python.org/guiqwt/_images/sift3.png), NumPy array editor (example: http://packages.python.org/guiqwt/_images/sift2.png), dictionary editor, source code editor, etc. Description of key features with tasty screenshots can be found at: http://code.google.com/p/spyderlib/wiki/Features On Windows platforms, Spyder is also available as a stand-alone executable (don't forget to disable UAC on Vista/7). This all-in-one portable version is still experimental (for example, it does not embed sphinx -- meaning no rich text mode for the object inspector) but it should provide a working version of Spyder for Windows platforms without having to install anything else (except Python 2.x itself, of course). Don't forget to follow Spyder updates/news: * on the project website: http://code.google.com/p/spyderlib/ * and on our official blog: http://spyder-ide.blogspot.com/ Last, but not least, we welcome any contribution that helps making Spyder an efficient scientific development/computing environment. Join us to help creating your favourite environment! (http://code.google.com/p/spyderlib/wiki/NoteForContributors) Enjoy! -Pierre

1 0

Using dtype=object
by Nathan Faggian 16 Nov '11

16 Nov '11

Hi, I am interested in the use of numpy with native python objects, like so: In [91]: import collections In [92]: testContainer = collections.namedtuple('testContainer', 'att1 att2 att3') In [93]: test1 = testContainer(1, 2, 3) In [94]: test2 = testContainer(4, 5, 6) In [95]: test1 Out[95]: testContainer(att1=1, att2=2, att3=3) In [96]: test2 Out[96]: testContainer(att1=4, att2=5, att3=6) In [97]: x = np.empty((2,2), dtype=object) In [98]: x[0,0] = test1 In [99]: x[1,1] = test2 In [100]: x Out[100]: array([[testContainer(att1=1, att2=2, att3=3), None], [None, testContainer(att1=4, att2=5, att3=6)]], dtype=object) Does anyone know if it possible to form a mask using the attributes of the objects stored in the ndarray? After a few failed attempts I am left wondering if I should use a prepared dtype instead. Cheers, Nathan.

2 1

Memory hungry reduce ops in Numpy
by Andreas Müller 15 Nov '11

15 Nov '11

Hi everybody. When I did some normalization using numpy, I noticed that numpy.std uses more ram than I was expecting. A quick google search gave me this: http://luispedro.org/software/ncreduce The site claims that std and other reduce operations are implemented naively with many temporaries. Is that true? And if so, is there a particular reason for that? This issues seems quite easy to fix. In particular the link I gave above provides code. Cheers, Andy

6 12

dedicated function for resize with averaging or rebin 2d arrays?
by Andrea Zonca 15 Nov '11

15 Nov '11

hi, I work in astrophysics where the most common programming language is currently IDL. A common request of people switching from IDL to python is the implementation of the REBIN function, which either downsizes a 2d array by averaging or increases its dimension by repeating its elements. In both cases the new shape must be an integer factor of the old shape. I believe it is a very handy function for quick smoothing of 2 dimensional data. I found a discussion about this topic in the archives: http://thread.gmane.org/gmane.comp.python.numeric.general/885/focus=894 Do you think it would be useful to add such function to numpy? I created a simple implementation to help in the discussion: https://gist.github.com/1348792 thanks, Andrea Zonca

4 6

numpy.int32 is not subclass of int, but numpy.int64 is
by MACKEITH Andrew 15 Nov '11

15 Nov '11

Could someone explain this? An instance of numpy.int32 is not an instance of int or numpy.int. An instance of numpy.int64 is an instance of int and numpy.int. I don't know if it is a bug in my linux build. Andrew >python26 Python 2.6.2 (r262:71600, Jul 8 2010, 11:49:56) [GCC 4.1.2 20070115 (SUSE Linux)] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> >>> import numpy >>> a = numpy.array((1,2,3)) >>> type(a[0]) <type 'numpy.int64'> >>> isinstance(a[0], int) True >>> isinstance(a[0], numpy.int) True >>> seven = numpy.int64(7) >>> isinstance(seven, numpy.int) True >>> isinstance(seven, int) True >>> >>> five = numpy.int32(5) >>> isinstance(five, numpy.int) False >>> isinstance(five, int) False >>> five 5 >>> seven 7 >>> issubclass(numpy.int64, numpy.int) True >>> issubclass(numpy.int32, numpy.int) False >>> >>> import sys >>> sys.maxint 9223372036854775807 >>> import platform >>> print platform.platform() Linux-2.6.32.12-0.7-default-x86_64-with-SuSE-11-x86_64 >>> This email and any attachments are intended solely for the use of the individual or entity to whom it is addressed and may be confidential and/or privileged. If you are not one of the named recipients or have received this email in error, (i) you should not read, disclose, or copy it, (ii) please notify sender of your receipt by reply email and delete this email and all attachments, (iii) Dassault Systemes does not accept or assume any liability or responsibility for any use of or reliance on this email. For other languages, go to http://www.3ds.com/terms/email-disclaimer

3 5

Odd-looking long double on windows 32 bit
by Matthew Brett 15 Nov '11

15 Nov '11

Hi, Sorry for my continued confusion here. This is numpy 1.6.1 on windows XP 32 bit. In [2]: np.finfo(np.float96).nmant Out[2]: 52 In [3]: np.finfo(np.float96).nexp Out[3]: 15 In [4]: np.finfo(np.float64).nmant Out[4]: 52 In [5]: np.finfo(np.float64).nexp Out[5]: 11 If there are 52 bits of precision, 2**53+1 should not be representable, and sure enough: In [6]: np.float96(2**53)+1 Out[6]: 9007199254740992.0 In [7]: np.float64(2**53)+1 Out[7]: 9007199254740992.0 If the nexp is right, the max should be higher for the float96 type: In [9]: np.finfo(np.float64).max Out[9]: 1.7976931348623157e+308 In [10]: np.finfo(np.float96).max Out[10]: 1.#INF I see that long double in C is 12 bytes wide, and double is the usual 8 bytes. So - now I am not sure what this float96 is. I was expecting 80 bit extended precision, but it doesn't look right for that... Does anyone know what representation this is? Thanks a lot, Matthew

3 11