Mailman 3 February 2009 - SciPy-Dev

scipy.spatial 'output' args
by David Warde-Farley Feb. 22, 2009

Feb. 22, 2009

Howdy, This is mostly a question for Damian (sorry, I seem to be bugging you on a lot of fronts!). I'm wondering if there's a principled reason why the pdist/squareform/etc. functions don't allow you to specify an output array. It seems hard to justify not having a way of avoiding repeated memory allocations if you're doing this more than once (as points change, for example -- if you're curious I'm implementing a 2-D embedding algorithm and the objective function relies on a lot of distances being recomputed). My guess is this is a MATLABism and that this was just an oversight. Is there a reason I shouldn't go about trying to add it? Would other people find this useful? David

2 1

on-line lectures and new text book using python, numpy, scipy
by John Stachurski Feb. 21, 2009

Feb. 21, 2009

Hi all, Thanks to all scipy and numpy developers for creating a great package. I'm an economist researching computational techniques, and I'm trying to wean economists off MATLAB and onto python/numpy/scipy. I have a book just published through MIT Press on computational economics using python: http://johnstachurski.net/book/book.html I've also written a fairly comprehensive set of lectures on python/numpy/ scipy with applications in economics http://johnstachurski.net/lectures/index.html I've added them to this page: http://www.scipy.org/Documentation listed under "Other". I hope people find them useful. All feedback is most welcome. Regards, John. -- John Stachurski Born to fish, forced to work Visit me at http://johnstachurski.net

3 3

RFR: #794 - linalg.eig segmentation fault on unpickled arrays
by Stéfan van der Walt Feb. 21, 2009

Feb. 21, 2009

Hi, Please review the patch attached to http://scipy.org/scipy/scipy/ticket/794 I'd like to know whether checking memory alignment and Fortran contiguity is sufficient. Please also let me know if you can think of a better way to avoid the LAPACK segfaults. Regards Stéfan

1 0

Matlab io bug; request for advice
by Matthew Brett Feb. 20, 2009

Feb. 20, 2009

Hi, I've just found another bug in the matlab io (mine of course). It's easy to fix, but I'd like some advice about a change in undocumented and somewhat unexpected behavior of the mat file writing. The bug is, that when writing an array to a matlab matrix variable, I had been appending 0s to the shape, for less than 2d arrays. For 0d arrays (scalars), this results in shape (0,0), which is read in matlab as an empty array. That's obviously a bug, and easy to fix. However, when passing in 1d arrays, it results in shapes like (3,0). Although this shape is illegal in matlab, it seems to kindly just change the 0 to a 1, resulting in a column vector. However, matlab tends to think of an unshaped vector as being a row vector: >> a = 1:12; >> size(a) ans = 1 12 And numpy thinks the same: >>> print np.atleast_2d(np.array([1,2])).shape (1, 2) It seems to me then, that I should assume the same, that a 1 dimensional array is a row vector. However, this will change the matlab shape of a 1 d array passed into the matlab mat file routines from a column vector to a row vector. Do y'all think I should: a) Given this is undocumented anyway, just switch to the row vector b) Keep 1d arrays as column vectors c) Raise a warning and change for a future release ? Thanks, Matthew

7 17

bug in scipy.ndimage.interpolation.zoom ?
by Nicolas Rougier Feb. 20, 2009

Feb. 20, 2009

Hello, >From the following script: import numpy from scipy.ndimage.interpolation import zoom n = 5 Z = numpy.ones((n)) print Z print zoom(Z,1,order=0) print zoom(Z.reshape((1,n)),(1,1),order=0) I get the following result: [ 1. 1. 1. 1. 1.] [ 1. 1. 1. 1. 1.] [[ 0. 0. 0. 0. 0.]] Is that the expected behavior ? Nicolas

2 2

[OT] Skipping a function wronlgy identified by nose
by Gael Varoquaux Feb. 19, 2009

Feb. 19, 2009

Sorry, this is off topic, but I know there are many nose users on this mailing list. Here is the problem: in our library, we have a perfectly valid function that is called 'onesample_test'. Due to the name, nose identifies it as a function (also because it is defined in a file called 'statistical_test.py'). Does anybody have an idea of what the right way to tell nose to skip it is? Nose is passing it the wrong number of arguments, and as a result it is appearing as a bogus error in the test suite. Thanks, Gaël

4 6

old link in releas notes 0.7.0
by Pieter Cristiaan de Groot Feb. 19, 2009

Feb. 19, 2009

Hello, I think that there is an old link in the release notes of 0.7.0. It seems that nose has moved from: http://code.google.com/p/python-nose/ to here: http://somethingaboutorange.com/mrl/projects/nose/ best, Pieter

4 6

Server errors
by Stéfan van der Walt Feb. 19, 2009

Feb. 19, 2009

Hi, Those "500 Internal Server Errors" are back again. Could someone with access to the machine please restart the necessary services? Thanks Stéfan

1 0

doc typo in fclusterdata() function
by Damian Eads Feb. 19, 2009

Feb. 19, 2009

Hi Martin, Good catch, thanks. I would appreciate it if you fixed the doc. I've forwarded your message to the scipy-dev mailing list. Create an account, reply with your account name, and someone will grant you editing rights. Thanks again, Damian On Wed, Feb 18, 2009 at 2:28 PM, Martin Spacek <mspacek(a)mm.st> wrote: > Hi Damian, > > The first entry in the arguments list in the docstring for > scipy.cluster.hierarchy.fclusterdata() is: > > - Z : ndarray > The hierarchical clustering encoded with the matrix returned > by the ``linkage`` function. > > It should be something like: > > - X : ndarray > ``n`` by ``m`` data matrix with ``n`` observations in ``m`` > dimensions > > This is in scipy 0.7rc2, the online scipy docs, and also in hcluster 0.2. > > I should get myself editing rights for the scipy docs :) > > Cheers, > > Martin

1 0

docs editing permission request for: mspacek
by Martin Spacek Feb. 18, 2009

Feb. 18, 2009

Hi, Could you please give me editing rights for the scipy/numpy docs at docs.scipy.org? My username is mspacek Cheers, Martin

2 1