Mailman 3 June 2007 - SciPy-User

Re: [SciPy-user] help importing integrate
by Trevis Crane June 11, 2007

June 11, 2007

> The toplevel scipy package doesn't import subpackages, so you need to do > > import scipy.integrate [Trevis Crane] I was confused because when you go to this link: http://www.scipy.org/doc/api_docs/ and click on the [+] button next to scipy under the modules heading I find an 'integrate' heading. Interestingly, if you click on the same button next to the scipy heading that is under the Packages heading, you again see 'integrate' listed. I assumed that these two lists contain the subpackages of scipy (under the packages heading) and the modules of scipy under the modules heading. Anyway, thanks for the clarification, trevis > > > -- > |>|\/|< > /----------------------------------------------------------------------- ---\ > |David M. Cooke http://arbutus.physics.mcmaster.ca/dmc/ > |cookedm(a)physics.mcmaster.ca > _______________________________________________ > SciPy-user mailing list > SciPy-user(a)scipy.org > http://projects.scipy.org/mailman/listinfo/scipy-user

1 0

How to solve name mangling ?
by Stef Mientki June 11, 2007

June 11, 2007

hello, the beauty of Python is that you can "rename" everything . In the languages I've been using up to now, an integer is an integer and stays an integer forever (has it's beauty too). The modules I write for myself, I always start with the "dangerous": form scipy import * Now if I take modules form others, or take code snippets (with forgetting to use the scipy import), I get different type of arrays, with all kind of weird behavior. The module at it's own works perfect, but if I call the module from another program, it doesn't work (as expected) anymore. Any clever solution to solve this without thinking about those tiny details ? Is using form scipy import * as the last global import in each module, a solution ? thanks, Stef Mientki

11 17

Building SciPy under SUSE 10.2 using Intel icc 10 and ifort 10
by rex June 10, 2007

June 10, 2007

I've managed to build NumPy 1.0.3 from source using the Intel icc compiler and the Intel MKL, and now I'm trying for SciPy. Reading the archives looking for help I found this: ==================================================================== Daniel Nogradi nogradi@gmail.... Sat May 19 12:03:49 CDT 2007 > > I have the following installed on a Fedora 3 box: [...] > > and am trying to do a fresh install of scipy-0.5.2 but "python > > setup.py build_ext" is complaining about a missing f95 executable and > > ld about missing dfftpack ..... > Pearu Peterson wrote: > You can ignore messages about missing f95 executable, they are just > parts of compiler detection tools. > > Libraries dfftpack and others are built with build_clib command, > so sole build_ext is not enough. Use > > python build > > to build scipy. Thanks very much, indeed, build_clib, build_ext, build works (in this order). ====================================================================== LAPACK and BLAS built successfully with the Intel compilers (details on SciPy-dev) using modifications of Steve Baum's instructions at: http://pong.tamu.edu/tiki/tiki-view_blog_post.php?blogId=6&postId=97 But every combination of options I can think of for the CL build of SciPy (recent svn) fails with one of two error messages: python setup.py build_clib --compiler=intel --fcompiler=intel build_ext --compiler=intel --fcompiler=intel build >inst12 [...] error: file 'dfftpack/*.f' does not exist python setup.py build_clib --compiler=intel --fcompiler=intel [...] error: file 'dfftpack/*.f' does not exist python setup.py build --compiler=intel --fcompiler=intel [...] error: '_fftpackmodule.c' missing Any help appreciated, thanks, -rex

1 0

fread (from numpyio) and stderr redirection...
by fred June 10, 2007

June 10, 2007

Hi, I want to test that data read from file with fread has exactly the right dimensions. - if less bytes are read that the required number of bytes, how can it be detected ? - if more bytes are tried to read, fread returns a warning message that I don't want to be displayed. How can I do this ? I have attached a sample file that shows the three cases: - data read from file has the right dimension: ok - data read from file has 1 byte instead of 2: no detection error - try to read from file 3 bytes instead of 2: warning message displayed Any suggestion ? Thanks in advance. -- http://scipy.org/FredericPetit

1 0

scipy.integrate and threading
by Eugen Wintersberger June 8, 2007

June 8, 2007

Hi there I have just started with using threads in python and want to use them together with the integration routines in scipy. However, there seems to be a serious problem with threadsafty. I want to run two threads which perform integration simultaneously. However, when I do so the program exits with a Segmentation Fault. If I use only one thread the program runs without problems. Is there any simple possibility to use scipy in threads? Tanks Eugen

2 1

indices, lists and arrays
by Emanuele Zattin June 6, 2007

June 6, 2007

i have something like this: sizes = ndimage.sum(bl_img, labels=bl_l, index=range(1,bl_n+1)) sizes = array(sizes) bl_obj_indices = where(sizes<21) bl_l[bl_objects[bl_obj_indices]] = 0 sizes was a list, but i converted it to an array in order to use the function where on it where returns an array of arrays, something like (array([14, 17]),) bl_objects is the output of ndimage.find_objects and is a list. this means that the assignment on the last row of the proposed code does not work. is there an elegant solution to solve the problem? list(bl_obj_indices) returns [array([14, 17])] so it does not do the trick. i would need something like [14, 17] any hint? Emanuele

2 3

weight of matrix or other object
by Robert VERGNES June 6, 2007

June 6, 2007

Hello, I have to stop calculation when my data variable reach a certain size ( weight). How can i know the weight in (Kb or Mb) of an object - either a list of matrix or a matrix itself ? Best Regards Robert --------------------------------- Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! Mail

2 2

scipy.odeint args question
by Luke June 6, 2007

June 6, 2007

I'm trying to write a tool that needs to be able to integrate n-dimensional systems of 1st order differential equations. I need to be able to basically plug various dynamical systems, along with their Jacobians, and any associated parameters that occur on the RHS of the ODE's that may need to be varied. My issue is with how scipy.odeint handles extra function arguments. My differential equations are of the form: def f(x,t,param): ... return dxdt def J(x,t,param): ... return dfdx The param argument would be a rank 1 array (or list), that gets used in the function definitions of the RHS. For example, the Lorenz equations have three parameters, sigma, r, and b, but other systems have other numbers of parameters, so it makes sense to just pass this as a vector (or list). If it is not done in this fashion, then systems with different numbers of parameters have to be hard-coded... really annoying. I would like to be able to call scipy.odeint something like: y = scipy.odeint(f, x0, t, args = param, Dfun = J) or y = scipy.odeint(f, x0, t, args = (param), Dfun = J) This is where I can't get things to work -- odeint needs a tuple for the args argument, and I can't figure out how to make it work. The following works, but is way to restrictive for what I need because I want to be able to make my code modular enough to be able to integrate *any* dynamical system that has an arbritrary number of parameters. Here is what works, for a system with three parameters, a, b, and c; def f(x,t,a,b,c): ... return dxdt def J(x,t,a,b,c): ... return dfdx y = scipy.odeint(f, x0, t, args = (a,b,c), Dfun = J) This above works fine, but again, if you need to evaluate f anywhere, then you have to know how many parameters it takes and call the function in a fashion that explicitly lays out how each parameter gets passed. Am I overlooking something really simple here that would make this work? I know in matlab's ode45 you can just pass scalars or arrays or matrices of additional parameters and it doesn't really matter -- you just pass them through. Thanks, ~Luke

3 2

f2py and ifort flags...
by fred June 5, 2007

June 5, 2007

Hi, 1) To compile my fortran code, I use the command line: ifort call_func.f -limf -Wl,--rpath -Wl,/usr/local/share/intel/fc/9.1.045/lib funcs.o -o call_func_f which works fine. 2) I want to use the same trick with f2py. So, from f2py manpage, my Makefile looks like: IFCVER = 9.1.045 PATH = /bin:/usr/bin:/usr/local/share/intel/fc/$(IFCVER)/bin LD_LIBRARY_PATH_INTEL = /usr/local/share/intel/fc/$(IFCVER)/lib LD_LIBRARY_PATH = /usr/local/lib:${LD_LIBRARY_PATH_INTEL} LDFLAGS = -limf -Wl,--rpath -Wl,$(LD_LIBRARY_PATH_INTEL) OPTIMIZATION = -O3 -xT -ipo all: funcs.o calc_KM.so calc_KV.so calc_output.so funcs.o: funcs.c gcc -Wall funcs.c -c -o funcs.o calc_KM.so: calc_KM.f funcs.o f2py --fcompiler=intel --opt="$(OPTIMIZATION)" --f90flags="$(LDFLAGS)" -c calc_KM.f funcs.o -m calc_KM .../... but this has no effect, ie intel libs are not linked. What am I doing wrong ? Cheers, -- http://scipy.org/FredericPetit

4 21

SciPy Journal
by Travis Oliphant June 4, 2007

June 4, 2007

Hi everybody, I'm sorry for the cross posting, but I wanted to reach a wide audience and I know not everybody subscribes to all the lists. I've been thinking more about the "SciPy Journal" that we discussed before and I have some thoughts. 1) I'd like to get it going so that we can push out an electronic issue after the SciPy conference (in September) 2) I think it's scope should be limited to papers that describe algorithms and code that are in NumPy / SciPy / SciKits. Perhaps we could also accept papers that describe code that depends on NumPy / SciPy that is also easily available. 3) I'd like to make a requirement for inclusion of new code in SciPy that it have an associated journal article describing the algorithms, design approach, etc. I don't see this journal article as being user-interface documentation for the code. I see this is as a place to describe why the code is organized as it is and to detail any algorithms that are used. 4) The purpose of the journal as I see it is to a) provide someplace to document what is actually done in SciPy and related software. b) provide a teaching tool of numerical methods with actual "people use-it" code that would be useful to researchers, students, and professionals. c) hopefully clever new algorithms will be developed for SciPy by people using Python that could be show-cased here d) provide a peer-review publication opportunity for people who contribute to open-source software 5) We obviously need associate editors and people willing to review submitted articles as well as people willing to submit articles. I have two articles that can be submitted within the next two months. What do other people have? As an example of the kind of thing a SciPy Journal would be useful for. I have recently over-hauled the interpolation.py file for SciPy by incorporating the B-spline stuff that is partly in fitpack. In the process I noticed two things: 1) I have (what seems to me) a different recursive algorithm for calculating derivatives of B-splines than I could find in fitpack. 2) I have developed a different way to determine the K-1 extra degrees of freedom for Kth-order spline fitting than I have seen before. The SciPy Journal would be a great place to document both of these things while describing the spline interpolation design of scipy.interpolate It is true that I could submit this stuff to other journals, but it seems like that doing that makes the information harder to find in the future and not easier. I'm also dissatisfied with how information exclusionary academic journals seem to be. They are catching up, but they are still not as accessible as other things available on the internet. Given the open nature of most scientific research, it is remarkable that getting access to the information is not as easy as it should be with modern search engines (if your internet domain does not subscribe to the e-journal). Comments and feedback is welcome. -Travis

8 9