Mailman 3 February 2006 - SciPy-User

object members
by Martin Wiechert Feb. 27, 2006

Feb. 27, 2006

Hi developers, any plans on when object members will be back? a = ndarray (shape = (10,), dtype = {'names': ['x'], 'formats': ['|O4']}) TypeError: fields with object members not yet supported. in numpy 0.9.5. Of course, this is better than the segfault in 0.9.4, but it would be quite inconvenient for my project to not have object members. My C code still produces dtypes with object members. Can I safely use them, as long as I make sure new arrays are properly initialised? Thanks, Martin.

1 0

Sparse matrix file format converter
by Nils Wagner Feb. 27, 2006

Feb. 27, 2006

I thought it might be of interest... From: "Mark Hoemmen" <mark.hoemmen(a)gmail.com> Date: Fri, 24 Feb 2006 14:38:06 -0800 Subject: Sparse matrix file format converter As part of my work for the UC Berkeley Benchmarking and Optimization Group ("BeBOP"), I've written the BeBOP Sparse Matrix Converter, a library and standalone utility for converting between Harwell-Boeing, MatrixMarket and Matlab (ASCII) sparse matrix file formats. I am releasing the library under a BSD license. Please try it out, and feel free to send me suggestions at mhoemmen(a)cs.berkeley.edu. Here is the website for the converter: http://www.cs.berkeley.edu/~mhoemmen/bebop/smc.html This library was put together with help from Richard Vuduc, Rajesh Nishtala, Ankit Jain, and other BeBOP colleagues. Mark Hoemmen mhoemmen(a)cs.berkeley.edu

1 0

Slow medians ?
by pgmdevlist＠mailcan.com Feb. 27, 2006

Feb. 27, 2006

Folks, I'm trying to implement by brute force some median-based linear regression techinques, requiring a lot of median computations (for n elements, I have about n*n medians to compute). And as I have a lot of data to process that way, I'm a bit pecky about optimization. I've noticed that for 1D and 2D arrays, the basic numpy version of median is far faster than the nD version of scipy.stats (you can still choose to compute along any of the 2 axes with numpy.median by playing with transpose). Would it be useful to modify the scipy.stats version, or should it be left to the user ? And on related matter, numpy based this time: the numpy.median version seems a bit strange: what's the idea behind using `msort` twice (around line 810 in the 0.9.5 version) ? Thanks for you feedback in advance -- Pierre GM

1 0

Re: [SciPy-user] scipy.org site down?
by Travis N. Vaught Feb. 26, 2006

Feb. 26, 2006

Looking into it--thanks for pointing this out. -----Original Message----- From: oliver.tomic(a)matforsk.no Date: Sunday, Feb 26, 2006 6:39 am Subject: Re: [SciPy-user] scipy.org site down? Same thing here. OLI scipy-user-bounces(a)scipy.net wrote on 26.02.2006 13:36:04: > It seems www.scipy.org and svn.scipy.org are both down again. They are replying to pings, but I get a 502 Bad Gateway error when trying to access www.scipy.org. Can anyone confirm this? > -- Ed > _______________________________________________ SciPy-user mailing list SciPy-user(a)scipy.net http://www.scipy.net/mailman/listinfo/scipy-user _______________________________________________ SciPy-user mailing list SciPy-user(a)scipy.net http://www.scipy.net/mailman/listinfo/scipy-user

1 1

Re: [SciPy-user] SciPy-user Digest, Vol 30, Issue 43
by Nick Fotopoulos Feb. 26, 2006

Feb. 26, 2006

Confirmed. Nick On Feb 26, 2006, at 6:35 AM, scipy-user-request(a)scipy.net wrote: > Date: Sun, 26 Feb 2006 13:36:04 +0100 > From: Ed Schofield <schofield(a)ftw.at> > Subject: [SciPy-user] scipy.org site down? > To: SciPy Users List <scipy-user(a)scipy.net> > Message-ID: <4401A0B4.7020208(a)ftw.at> > Content-Type: text/plain; charset=UTF-8 > > > It seems www.scipy.org and svn.scipy.org are both down again. They > are > replying to pings, but I get a 502 Bad Gateway error when trying to > access www.scipy.org. Can anyone confirm this? > > -- Ed

2 1

scipy.org site down?
by Ed Schofield Feb. 26, 2006

Feb. 26, 2006

It seems www.scipy.org and svn.scipy.org are both down again. They are replying to pings, but I get a 502 Bad Gateway error when trying to access www.scipy.org. Can anyone confirm this? -- Ed

2 1

D1MACH not set
by Christian Kristukat Feb. 26, 2006

Feb. 26, 2006

Hi, using a recent version of numpy/scipy from svn I get the following error when calling any of the optimize.fmin_* minimizers. tnc: Version 1.2, (c) 2002-2003, Jean-Sebastien Roy (js(a)jeannot.org) tnc: RCS ID: @(#) $Jeannot: tnc.c,v 1.201 2004/04/02 22:36:25 js Exp $ Adjust D1MACH by uncommenting data statements appropriate for your machine. STOP 779 In [2]: numpy.__version__ Out[2]: '0.9.6.2168' In [4]: scipy.__version__ Out[4]: '0.4.7.1620' Regards, Christian

1 0

Messing with missing values
by pgmdevlist＠mailcan.com Feb. 25, 2006

Feb. 25, 2006

Folks, Most of the data I work with have missing values, and I rely on 'ma' a lot. Unfortunately, most of the numpy/scipy functions don't handle missing values nicely. Not a problem I thought, I just have to adapt the functions I need. I have 2 options: wrapping the initial functions in tests, or recode the initial functions. * According to your experience, which is the most efficient way to go ? * I have a function that outputs either a float or MA.masked. I call this function recursively, appending the result to a list, and then trying to process the list as an array. How could I mask the values corresponding to MA.masked in the final list, without having to check every single element? Thanks for your ideas -- Pierre GERARD-MARCHANT Driftmier Engineering Center UGA Athens, GA 30602 -- "It's more fun to compute" - Kraftwerk

1 0

Wrapping geqp3 (Rank revealing QR)
by Nils Wagner Feb. 25, 2006

Feb. 25, 2006

Hi all, I was trying to wrap geqp3...(my homework as suggested by Arnd ;-) ) It turns out that it is not straightforward (at least for me). I have used an example taken from generic_flapack.pyf subroutine <tchar=s,d,c,z>geqrf(m,n,a,tau,work,lwork,info) ! qr_a,tau,work,info = geqrf(a,lwork=3*n,overwrite_a=0) ! Compute a QR factorization of a real M-by-N matrix A: ! A = Q * R. callstatement (*f2py_func)(&m,&n,a,&m,tau,work,&lwork,&info) callprotoargument int*,int*,<type_in_c>*,int*,<type_in_c>*,<type_in_c>*,int*,int* integer intent(hide),depend(a):: m = shape(a,0) integer intent(hide),depend(a):: n = shape(a,1) <type_in> dimension(m,n),intent(in,out,copy,out=qr) :: a <type_in> dimension(MIN(m,n)),intent(out) :: tau integer optional,intent(in),depend(n),check(lwork>=n||lwork==-1) :: lwork=3*n <type_in> dimension(MAX(lwork,1)),intent(out),depend(lwork) :: work integer intent(out) :: info end subroutine <tchar=s,d,c,z>geqrf Some observations: The number of arguments differ wr.t. to complex (C,Z) and real matrices (S,D). How is this handled with scipy ? SUBROUTINE CGEQP3( M, N, A, LDA, JPVT, TAU, WORK, LWORK, RWORK, $ INFO ) SUBROUTINE DGEQP3( M, N, A, LDA, JPVT, TAU, WORK, LWORK, INFO ) Here is my first (incomplete !!) attempt to adapt the wrapper . How can I proceed ? ! -*- f90 -*- ! Note: the context of this file is case sensitive. python module wrap_lap ! in interface ! in :wrap_lap subroutine <tchar=s,d>geqp3(m,n,a,lda,jpvt,tau,work,lwork,info) ! in :wrap_lap:dgeqp3.f ! rrqr_a,tau,work,info = geqp3(a,lwork=...,overwrite_a=0) ! Compute the a QR factorization with column pivoting of a matrix A ! A*P = Q*R using Level 3 BLAS callstatement (*f2py_func)(&m,&n,a,&m,tau,work,&lwork,&info) callprotoargument int*,int*,<type_in_c>*,int*,<type_in_c>*,int*,int* integer intent(hide),depend(a):: m=shape(a,0) integer intent(hide),depend(a):: n=shape(a,1) <type_in> dimension(m,n), intent(in,out,copy,out=rrqr) :: a <type_in> dimension(MIN(m,n)),intent(out) :: tau integer dimension(*) :: jpvt double precision dimension(*), intent(out) :: tau double precision dimension(*), intent(out) :: work <type_in> dimension(lwork),intent(hide,cache),depend(lwork):: work integer intent(out):: info end subroutine <tchar=s,d>geqp3 end interface end python module wrap_lap ! This file was auto-generated with f2py (version:2_2163). ! See http://cens.ioc.ee/projects/f2py2e/ ~ Nils

2 8

Re: [SciPy-user] in-place data permutation
by Brian Hawthorne Feb. 25, 2006

Feb. 25, 2006

Hello, I'm looking for an efficient implementation of permuting the positions of data in an array which doesn't require copying the array. Specifically, I am shifting an array by half its length along the 0th dimension, in order to put "textbook" ordered frequency data into the packed format expected by inverse_fft. I'm surprised this is not already implemented in numpy... seems like it would be a common need. Am I missing something obvious? Brian Hawthorne UC Berkeley Brain Imaging Center

2 1