Mailman 3 February 2014 - SciPy-Dev

Reordering methods for Sparse matrices
by Paul Nation 22 Feb '14

22 Feb '14

I have written two sparse matrix reordering functions in Cython and I am wondering if they would be worthy additions to the scipy.sparse module? These are currently being used in our QuTiP library (qutip.org). The first function finds the Reverse Cuthill-Mckee (RCM) ordering [1] of a symmetric sparse matrix, similar to the Matlab ‘symrcm' function. This is useful for reducing the bandwidth of the underlying matrix to help eliminate fill-in when using direct or iterative solvers. If the matrix is not symmetric, it looks at A+trans(A). Since the matrix is symmetric, this works for both CSR and CSC matrices. The second, is a maximal traversal algorithm based on a breadth-first-search method [2]. This returns a set of row permutations that makes the diagonal zero-free. Since the permutation is not symmetric, this works only on CSC matrices. Like RCM this is useful for preordering in both direct and iterative solvers. [1] E. Cuthill and J. McKee, "Reducing the Bandwidth of Sparse Symmetric Matrices”, ACM '69 Proceedings of the 1969 24th national conference, (1969). [2] I. S. Duff, K. Kaya, and B. Ucar, "Design, Implementation, and Analysis of Maximum Transversal Algorithms", ACM Trans. Math. Softw. 38, no. 2, (2011). Best regards, Paul

2 1

"Anonymous" bug reporting: useful?
by Andreas Hilboll 22 Feb '14

22 Feb '14

I was wondering if it would be good if users could report bugs without having to sign up for a Github account. From a technical viewpoint, this should be possible, using a dedicated email account, some server-side scripting, and the Github API. Do you have any feeling for how many bugs do not get reported because a user doesn't want to go through the hassle of creating an account just to report a bug? I have the feeling that it must be as simple as possible to report bugs. And having to create an account (be it at Github or by signing up for a mailing list) is an obstacle to that. Any thoughts / ideas? -- Andreas.

4 6

The scipy.optimize.minimize() documentation: highlight show_options?
by Yuxiang Wang 21 Feb '14

21 Feb '14

Dear all, I have a feel that the "show_options" is not as highlighted as in the documentation of the scipy.optimize.minimize(), as I personally took quite a while to see it. I now know how to use it, but I was wondering whether anyone echoes with me in the following suggestions: 1) Place the show_options('minimize', method) in the "See also:", so it is more highlighted. 2) Display scipy.optimize.show_options(), instead of show_options() because the latter confused me... well, that could also just be me :( 3) This is one that I haven't thought through yet: I don't think most people will first start reading the parameter arguments to the end and see the "show_options" description. Rather, one tends to start reading the notes about the algorithm that they care about. Could we add a list (only the parameter names) in the notes following each method? It could be simple, like: "Method L-BFGS-B uses the L-BFGS-B algorithm [R98], [R99] for bound constrained minimization. Available options are ftol, gtol, maxcor and maxiter. Please use scipy.optimize.show_options('minimize', 'L-BFGS-B') for details." Please let me know how you may think. I may highly be wrong, but I'd like to bring this topic up, in case this issue get slipped. Thanks! -Shawn -- Yuxiang "Shawn" Wang Gerling Research Lab University of Virginia yw5aj(a)virginia.edu +1 (434) 284-0836 https://sites.google.com/a/virginia.edu/yw5aj/

1 0

Fixing the scipy.optimize.show_options
by Yuxiang Wang 21 Feb '14

21 Feb '14

Dear all, I found this function missing the option 'eps' when I checked L-BFGS-B algorithm. This is what I got. In [1]: from scipy.optimize import show_options In [2]: show_options('minimize', 'l-bfgs-b') L-BFGS-B options: ftol : float The iteration stops when ``(f^k - f^{k+1})/max{|f^k|,|f^{k+1}|,1} <= ftol``. gtol : float The iteration will stop when ``max{|proj g_i | i = 1, ..., n} <= gtol`` where ``pg_i`` is the i-th component of the projected gradient. maxcor : int The maximum number of variable metric corrections used to define the limited memory matrix. (The limited memory BFGS method does not store the full hessian but uses this many terms in an approximation to it.) maxiter : int Maximum number of function evaluations. -Shawn -- Yuxiang "Shawn" Wang Gerling Research Lab University of Virginia yw5aj(a)virginia.edu +1 (434) 284-0836 https://sites.google.com/a/virginia.edu/yw5aj/

1 0

Maximum length sequence
by Eric Larson 21 Feb '14

21 Feb '14

Maximum length sequences (MLS) are useful in signal processing for finding impulse responses. I can't find a great implementation of MLS in Python, but maybe I've missed one somewhere. Would this be something good to put in scipy.signal, perhaps as scipy.signal.mls? Cheers, Eric

4 6

An inconsistency in scipy.optimize.minimize
by Yuxiang Wang 21 Feb '14

21 Feb '14

Dear all, I have been using the wrapper in scipy.optimize, the minimize(), to call L-BFGS-B method. One option that I'd like to set, is called "epsilon", as shown in this page. http://docs.scipy.org/doc/scipy/reference/generated/scipy.optimize.fmin_l_b… However, by digging into the code, I realize that this value's name is called "eps" instead of "epsilon" in the minimize wrapper. That is to say, options = {'epsilon': 1e-3} will not work and says it is not recgonized, but options = {'eps': 1e-3} can work. I think this will cause future users a lot of confusions (not everyone will dig into the code to find out what it is called). Or... it could be me understanding something wrong as well. What do you guys think? -Shawn -- Yuxiang "Shawn" Wang Gerling Research Lab University of Virginia yw5aj(a)virginia.edu +1 (434) 284-0836 https://sites.google.com/a/virginia.edu/yw5aj/

4 7

ENH: linprog function for linear programming
by Rob Falck 20 Feb '14

20 Feb '14

I've finally gotten around to putting a linear programming routine into a good state for inclusion into scipy. I've tested with somewhat large problems (400 variables, 40 constraints) and it solves within a second or two. Right now it's only uses a two-phase dense-matrix based simplex solver, but the overall "linprog" routine is intended to function in much the way that minimize does, where different methods can be used as they are added in the future. The pull request has been submitted and passes all tests. I welcome people to put it to the test and find any problems I may have missed. -- - Rob Falck

1 0

Matlab's fminbnd vs. scipy.optimize
by Petr Baudis 20 Feb '14

20 Feb '14

Hi! I'd like to ask if anyone knows of any important differences between Matlab's fminbnd and scipy.optimize.fminbound (minimize_scalar method Bounded)? Can they be assumed to behave and perform in a roughly equivalent way? Thanks, Petr "Pasky" Baudis

1 0

Pull request labels
by Pauli Virtanen 19 Feb '14

19 Feb '14

Hi, I'm going to tag some unfinished pull requests with a "needs-work" label. This should help with the problem that it's not easy to see from the list which ones are ready to go and which are waiting for someone to do something. Since the labels cannot be changed by the PR submitter, please shout in the comments if the label is out of date. Pauli

3 3

ANN: Scipy 0.13.3 release
by Ralf Gommers 16 Feb '14

16 Feb '14

Hi, I'm happy to announce the availability of the scipy 0.13.3 release. This is a bugfix only release; it contains fixes for regressions in ndimage and weave. Source tarballs can be found at https://sourceforge.net/projects/scipy/files/scipy/0.13.3/ and on PyPi. Release notes copied below, binaries will follow later (the regular build machine is not available for the next two weeks). Cheers, Ralf ========================== SciPy 0.13.3 Release Notes ========================== SciPy 0.13.3 is a bug-fix release with no new features compared to 0.13.2. Both the weave and the ndimage.label bugs were severe regressions in 0.13.0, hence this release. Issues fixed ------------ - 3148: fix a memory leak in ``ndimage.label``. - 3216: fix weave issue with too long file names for MSVC. Other changes ------------- - Update Sphinx theme used for html docs so ``>>>`` in examples can be toggled. Checksums ========= 0547c1f8e8afad4009cc9b5ef17a2d4d release/installers/scipy-0.13.3.tar.gz 20ff3a867cc5925ef1d654aed2ff7e88 release/installers/scipy-0.13.3.zip

1 1