Mailman 3 November 2012 - NumPy-Discussion

How to Keep An Array Two Dimensional
by Tom Bennett 26 Nov '12

26 Nov '12

Hi, I am trying to extract n columns from an 2D array and then operate on the extracted columns. Below is the code: A is an MxN 2D array. u = A[:,:n] #extract the first n columns from A B = np.dot(u, u.T) #take outer product. This code works when n>1. However, when n=1, u becomes an 1D array instead of an Mx1 2D array and the code breaks down. I wonder if there is any way to keep u=A[:,:n] an Mxn array no matter what value n takes. I do not want to use matrix because array is more convenient in other places. Thanks, Tom

3 3

adding rfftfreq
by Ralf Gommers 25 Nov '12

25 Nov '12

Hi, https://github.com/numpy/numpy/pull/484 adds an rfftfreq function to numpy.fft. Looks logical to me, but before merging I want to check anyway if no one objects to that. Ralf

2 1

Re: [Numpy-discussion] numpy where function on different sized arrays
by Siegfried Gonzi 25 Nov '12

25 Nov '12

On 25 Nov 2012, at 00:29, numpy-discussion-request(a)scipy.org wrote: > > Message: 3 > Date: Sat, 24 Nov 2012 23:23:36 +0100 > From: Da?id <davidmenhur(a)gmail.com> > Subject: Re: [Numpy-discussion] numpy where function on different > sized arrays > To: Discussion of Numerical Python <numpy-discussion(a)scipy.org> > Message-ID: > <CAJhcF=2nmVeeQUo5Q45OUcQK+7KnQ_u0tvzUyOSXGpQbLSUy+Q(a)mail.gmail.com> > Content-Type: text/plain; charset=ISO-8859-1 > > A pure Python approach could be: > > for i, x in enumerate(a): > for j, y in enumerate(x): > if y in b: > idx.append((i,j)) > > Of course, it is slow if the arrays are large, but it is very > readable, and probably very fast if cythonised. Thanks for all the answers. In that particular case speed is not important (A is 360x720 and b and c is lower than 10 in terms of dimension). However, I stumbled across similar comparison problems in IDL a couple of times where speed was crucial. My own solution or attempt was this: == def fu(A, b, c): for x, y in zip(b,c): indx = np.where(A == x) A[indx] = y return A == > > > The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.

1 0

numpy where function on different sized arrays
by Siegfried Gonzi 25 Nov '12

25 Nov '12

Hi all This must have been answered in the past but my google search capabilities are not the best. Given an array A say of dimension 40x60 and given another array/vector B of dimension 20 (the values in B occur only once). What I would like to do is the following which of course does not work (by the way doesn't work in IDL either): indx=where(A == B) I understand A and B are both of different dimensions. So my question: what would the fastest or proper way to accomplish this (I found a solution but think is rather awkward and not very scipy/numpy-tonic tough). Thanks -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.

3 4

Z-ordering (Morton ordering) for numpy
by Gamblin, Todd 25 Nov '12

25 Nov '12

Hi all, In the course of developing a network mapping tool I'm working on, I also developed some python code to do arbitrary-dimensional z-order (morton order) for ndarrays. The code is here: https://github.com/tgamblin/rubik/blob/master/rubik/zorder.py There is a function to put the elements of an array in Z order, and another one to enumerate an array's elements in Z order. There is also a ZEncoder class that can generate Z-codes for arbitrary dimensions and bit widths. I figure this is something that would be generally useful. Any interest in having this in numpy? If so, what should the interface look like and can you point me to a good spot in the code to add it? I was thinking it might make sense to have a Z-order iterator for ndarrays, kind of like ndarray.flat. i.e.: arr = np.empty([4,4], dtype=int) arr.flat = range(arr.size) for elt in arr.zorder: print elt, 0 4 1 5 8 12 9 13 2 6 3 7 10 14 11 15 Or an equivalent to ndindex: arr = np.empty(4,4, dtype=int) arr.flat = range(arr.size) for ix in np.zindex(arr.shape): print ix, (0, 0) (1, 0) (0, 1) (1, 1) (2, 0) (3, 0) (2, 1) (3, 1) (0, 2) (1, 2) (0, 3) (1, 3) (2, 2) (3, 2) (2, 3) (3, 3) Thoughts? -Todd ______________________________________________________________________ Todd Gamblin, tgamblin(a)llnl.gov, http://people.llnl.gov/gamblin2 CASC @ Lawrence Livermore National Laboratory, Livermore, CA, USA

4 4

unsubscribe
by Mogus Mochena 24 Nov '12

24 Nov '12

1 0

Re: [Numpy-discussion] numpy where function on different size
by Siegfried Gonzi 24 Nov '12

24 Nov '12

> Message: 6 > Date: Sat, 24 Nov 2012 20:36:45 +0000 > From: Siegfried Gonzi <sgonzi(a)staffmail.ed.ac.uk> > Subject: [Numpy-discussion] numpy where function on different size > Hi all >This must have been answered in the past but my google search > capabilities are not the best. >Given an array A say of dimension 40x60 and given another array/vector B > of dimension 20 (the values in B occur only once). >What I would like to do is the following which of course does not work > (by the way doesn't work in IDL either): >indx=where(A == B) >I understand A and B are both of different dimensions. So my question: > what would the fastest or proper way to accomplish this (I found a > solution but think is rather awkward and not very scipy/numpy-tonic > tough). I should clarify: where(A==B, C, A) C is of equal dimension than B. Basically: everywhere where A equals e.g. B[0] replace by C[0]; or where B[4] equals A replace very occurence in A by C[4]. -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.

1 0

Unexpected RuntimeWarning
by Bob Dowling 23 Nov '12

23 Nov '12

I have a simple function defined in the following snippet: --- start --- import numpy def chebyshev(x, m): '''Calculates Chebyshev functions of the first kind using the trigonometric identities.''' theta = numpy.where( numpy.abs(x)<=1.0, numpy.arccos(x), numpy.arccosh(numpy.abs(x)) ) y = numpy.where( numpy.abs(x)<=1.0, numpy.cos(m*theta), numpy.cosh(m*theta) * numpy.where( x > 0.0, 1.0, -1.0 )**m ) return y if __name__ == '__main__': x = numpy.linspace(-2.0, 2.0, 21) y = chebyshev(x,3) print(y) --- end --- I'm using the numpy.where() call to extract only legal values for the circular and hyperbolic trigonometric functions. But I still get warnings that I'm passing invalid anrguments: --- start--- $ python3 demo.py demo.py:8: RuntimeWarning: invalid value encountered in arccos numpy.arccos(x), demo.py:9: RuntimeWarning: invalid value encountered in arccosh numpy.arccosh(numpy.abs(x)) [ -2.60000000e+01 -1.79280000e+01 -1.15840000e+01 -6.77600000e+00 -3.31200000e+00 -1.00000000e+00 3.52000000e-01 9.36000000e-01 9.44000000e-01 5.68000000e-01 -1.83697020e-16 -5.68000000e-01 -9.44000000e-01 -9.36000000e-01 -3.52000000e-01 1.00000000e+00 3.31200000e+00 6.77600000e+00 1.15840000e+01 1.79280000e+01 2.60000000e+01] --- end --- (I get the same with Python 2 so don't get excited about that.) I'm guessing that numpy.where() is evaluating the complete arccos and arccosh arrays and therefore getting invalid arguments. Now, I can turn off the warnings with numpy.seterr(invalid='ignore') but that's not what I would regard as good practice. Is there a "numpythonic" way to address the issue?

3 3

When are 0-d arrays writeable?
by Charles R Harris 23 Nov '12

23 Nov '12

Examples, In [13]: ones(()).flags.writeable Out[13]: True In [14]: (-ones(())).flags.writeable Out[14]: False In [15]: (-1*ones(())).flags.writeable Out[15]: False In [16]: (1 + ones(())).flags.writeable Out[16]: False In [17]: array(1) Out[17]: array(1) In [18]: array(1).shape Out[18]: () In [19]: array(1).flags.writeable Out[19]: True Chuck

3 3

[ANN] Stochastic programming and optimization addon for FuncDesigner v. 0.421
by Dmitrey 23 Nov '12

23 Nov '12

hi all, I'm glad to inform you that stochastic programming and optimization addon for FuncDesigner v. 0.421 has been released. Now you can use gradient-based solvers for numerical optimization, such as ALGENCAN, IPOPT, ralg, gsubg etc. Usually they work faster than derivative-free (such as scipy_cobyla, BOBYQA) or global (GLP) solvers, e.g. on this example ALGENCAN time elapsed is less than 1 second while scipy_cobyla spend ~20 sec. However pay attention that having function P() in your problem may bring nonconvexity or some other issues to the solver optimization trajectory, thus sometimes you'll have to use derivative-free or GLP solvers (e.g. de) instead. FuncDesigner is free (BSD license) cross-platform Python language written software, while its stochastic programming and optimization addon, written by same authors, is free for small-scaled problems with educational or research purposes only. For more details visit our website http://openopt.org ------------------------------------- Regards, D. http://openopt.org/Dmitrey

1 0

NumPy-Discussion November 2012