[Numpy-discussion] how to use ldexp?
hi all, I have tried the example from numpy/add_newdocs.py np.ldexp(5., 2) but instead of the 20 declared there it yields TypeError: function not supported for these types, and can't coerce safely to supported types I have tried arrays but it yields same error np.ldexp(np.array([5., 2.]), np.array([2, 1])) Traceback (innermost last): File stdin, line 1, in module TypeError: function not supported for these types, and can't coerce safely to supported types So, how can I use ldexp? np.__version__ = '1.4.0.dev6972' Thank you in advance, D. ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
[Numpy-discussion] numpy ufuncs and COREPY - any info?
hi all, has anyone already tried to compare using an ordinary numpy ufunc vs that one from corepy, first of all I mean the project http://socghop.appspot.com/student_project/show/google/gsoc2009/python/t124024628235 It would be interesting to know what is speedup for (eg) vec ** 0.5 or (if it's possible - it isn't pure ufunc) numpy.dot(Matrix, vec). Or any another example. ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] how to use ldexp?
dmitrey wrote: hi all, I have tried the example from numpy/add_newdocs.py np.ldexp(5., 2) but instead of the 20 declared there it yields TypeError: function not supported for these types, and can't coerce safely to supported types Which OS/Compiler are you using ? David ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] how to use ldexp?
dmitrey wrote: I have updated numpy to latest '1.4.0.dev7008', but the bug still remains. I use KUBUNTU 9.04, compilers - gcc (using build-essential), gfortran. D. Can you post the build output (after having removed the build directory : rm -rf build python setup.py build build.log) ? David ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] how to use ldexp?
On May 21, 11:29 am, David Cournapeau da...@ar.media.kyoto-u.ac.jp wrote: dmitrey wrote: I have updated numpy to latest '1.4.0.dev7008', but the bug still remains. I use KUBUNTU 9.04, compilers - gcc (using build-essential), gfortran. D. Can you post the build output (after having removed the build directory : rm -rf build python setup.py build build.log) ? David ok, it's here http://pastebin.com/mb021e11 D. ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] how to use ldexp?
Thu, 21 May 2009 01:45:54 -0700, dmitrey wrote: I have updated numpy to latest '1.4.0.dev7008', but the bug still remains. I use KUBUNTU 9.04, compilers - gcc (using build-essential), gfortran. Worksforme on Ubuntu 9.04, on python2.6 and python2.5. Should be the same platform. -- Pauli Virtanen ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] how to use ldexp?
Thu, 21 May 2009 09:26:18 +, Pauli Virtanen wrote: Thu, 21 May 2009 01:45:54 -0700, dmitrey wrote: I have updated numpy to latest '1.4.0.dev7008', but the bug still remains. I use KUBUNTU 9.04, compilers - gcc (using build-essential), gfortran. Worksforme on Ubuntu 9.04, on python2.6 and python2.5. Should be the same platform. This was on 32-bit machine. I can reproduce this on a 64-bit platform, current SVN head: np.ldexp(5, 2) Traceback (most recent call last): File stdin, line 1, in module TypeError: function not supported for these types, and can't coerce safely to supported types np.ldexp(5, np.int32(2)) 20.0 np.ldexp.types ['fi-f', 'di-d', 'gi-g'] So for some reason the second argument tries to cast Python int to int64, and there's no loop to handle this. -- Pauli Virtanen ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
[Numpy-discussion] matrix default to column vector?
By default, it looks like a 1-dim ndarray gets converted to a row vector by the matrix constructor. This seems to lead to some odd behavior such as a[1] yielding the 2nd element as an ndarray and throwing an IndexError as a matrix. Is it possible to set a flag to make the default be a column vector? Thanks, Jason -- Jason Rennie Research Scientist, ITA Software http://www.itasoftware.com/ ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
[Numpy-discussion] where are the benefits of ldexp and/or array times 2?
Hi all, I expected to have some speedup via using ldexp or multiplying an array by a power of 2 (doesn't it have to perform a simple shift of mantissa?), but I don't see the one. Have I done something wrong? See the code below. from scipy import rand from numpy import dot, ones, zeros, array, ldexp from time import time N = 1500 A = rand(N, N) b = rand(N) b2 = 2*ones(A.shape, 'int32') I = 100 t = time() for i in xrange(I): dot(A, b) # N^2 multiplications + some sum operations #A * 2.1 # N^2 multiplications, so it should consume no greater than 1st line time #ldexp(A, b2) # it should consume no greater than prev line time, isn't it? print 'time elapsed:', time() - t # 1st case: 0.62811088562 # 2nd case: 2.00850605965 # 3rd case: 6.79027700424 # Let me also note - # 1) using b = 2 * ones(N) or b = zeros(N) doesn't yield any speedup vs b = rand() # 2) using A * 2.0 (or mere 2) instead of 2.1 doesn't yield any speedup, despite it is exact integer power of 2. ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] where are the benefits of ldexp and/or array times 2?
On Thu, May 21, 2009 at 10:26, dmitrey dmitrey.kros...@scipy.org wrote: Hi all, I expected to have some speedup via using ldexp or multiplying an array by a power of 2 (doesn't it have to perform a simple shift of mantissa?), Addition of the exponent, not shift of the mantissa. but I don't see the one. I said there *might* be a speedup, but it was probably going to be insignificant. The overhead of using frexp and ldexp probably outweighs any benefits. -- Robert Kern I have come to believe that the whole world is an enigma, a harmless enigma that is made terrible by our own mad attempt to interpret it as though it had an underlying truth. -- Umberto Eco ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] ANN: Cython 0.11.2 released
Where can I find release notes? (It would be helpful if I can point to a URL as part of the fedora release) OK, I put my email announcement up here: http://wiki.cython.org/ReleaseNotes-0.11.2 Tell me if you need something else (different format or level of detail -- the list of tickets on trac is always the most accurate thing). Dag Sverre ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
[Numpy-discussion] memoryerror with numpy.fromfile
I am getting a MemoryError from a numpy.fromfile() call in an application I am trying to deploy. Normally I would assume that this would mean that I don't have enough memory available on the system. However, if I run vmstat (Linux) at the same time as my process, I see that I have 3+ Gigabytes of memory free, and no swap space being used. I can't think of a way to track down this problem, so I'm punting to the list. The only thing I can imagine is that someone Python has been allocated X amount of space (very small relative to the memory actually available), and is asking for more than X. I don't know if this is true, or if there is even a way to check it. Sigh. Version info: Linux RedHat 5 kernel 2.6.18-128.1.10.el5PAE Python 2.5.4 -- EPD_Py25 4.3.0 numpy 1.3.0 Can anyone suggest more tests that I can do? Thanks, Mike ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] memoryerror with numpy.fromfile
Thu, 21 May 2009 10:31:28 -0600, Michael Hearne wrote: I am getting a MemoryError from a numpy.fromfile() call in an application I am trying to deploy. Normally I would assume that this would mean that I don't have enough memory available on the system. However, if I run vmstat (Linux) at the same time as my process, I see that I have 3+ Gigabytes of memory free, and no swap space being used. If you are on a 32-bit platform, the maximum addressable memory for a single process is limited to 3 GB, and what can be allocated can be less than this because of memory fragmentation. Also, you should check that you don't have an ulimit set for virtual/RSS memory. -- Pauli Virtanen ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] memoryerror with numpy.fromfile
On Thu, May 21, 2009 at 10:31 AM, Michael Hearne mhea...@usgs.gov wrote: I am getting a MemoryError from a numpy.fromfile() call in an application I am trying to deploy. Normally I would assume that this would mean that I don't have enough memory available on the system. However, if I run vmstat (Linux) at the same time as my process, I see that I have 3+ Gigabytes of memory free, and no swap space being used. I can't think of a way to track down this problem, so I'm punting to the list. The only thing I can imagine is that someone Python has been allocated X amount of space (very small relative to the memory actually available), and is asking for more than X. I don't know if this is true, or if there is even a way to check it. Sigh. Version info: Linux RedHat 5 kernel 2.6.18-128.1.10.el5PAE Python 2.5.4 -- EPD_Py25 4.3.0 numpy 1.3.0 Can anyone suggest more tests that I can do? How big is the file and what type are you importing to? Chuck ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] wiki page correction
It's an immutable page. Can someone who already has access make the edit? On Wed, May 20, 2009 at 2:36 PM, Pauli Virtanen p...@iki.fi wrote: Wed, 20 May 2009 12:08:46 -0700, Grant Kelly wrote: I believe there is an error on this wiki page: http://www.scipy.org/NumPy_for_Matlab_Users MATLAB y=x(2,:) PYTHON y = x[2,:].copy() shouldn't the Python version be: y = x[1,:].copy() If not, please advise. Yes, it should be x[1,:].copy(). Please feel free to correct it. -- Pauli Virtanen ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] memoryerror with numpy.fromfile
All: Never mind! The file I was attempting to read was part of an (apparently) silently incomplete download, and as such, the file size didn't match the metadata in the header file describing the data file, and I was reading beyond the end of the file. I would submit that MemoryError is perhaps a little misleading for this particular case, but oh well. Thanks for the comments! --Mike On May 21, 2009, at 11:46 AM, Charles R Harris wrote: On Thu, May 21, 2009 at 10:31 AM, Michael Hearne mhea...@usgs.gov wrote: I am getting a MemoryError from a numpy.fromfile() call in an application I am trying to deploy. Normally I would assume that this would mean that I don't have enough memory available on the system. However, if I run vmstat (Linux) at the same time as my process, I see that I have 3+ Gigabytes of memory free, and no swap space being used. I can't think of a way to track down this problem, so I'm punting to the list. The only thing I can imagine is that someone Python has been allocated X amount of space (very small relative to the memory actually available), and is asking for more than X. I don't know if this is true, or if there is even a way to check it. Sigh. Version info: Linux RedHat 5 kernel 2.6.18-128.1.10.el5PAE Python 2.5.4 -- EPD_Py25 4.3.0 numpy 1.3.0 Can anyone suggest more tests that I can do? How big is the file and what type are you importing to? Chuck ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] memoryerror with numpy.fromfile
Thu, 21 May 2009 12:14:18 -0600, Michael Hearne wrote: All: Never mind! The file I was attempting to read was part of an (apparently) silently incomplete download, and as such, the file size didn't match the metadata in the header file describing the data file, and I was reading beyond the end of the file. I would submit that MemoryError is perhaps a little misleading for this particular case, but oh well. Well, that it raises MemoryError in this case is a bug, so maybe a ticket should be filed. There's another bug associated with reading empty files: http://projects.scipy.org/numpy/ticket/1115 which is maybe related. -- Pauli Virtanen ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] wiki page correction
Thu, 21 May 2009 10:54:21 -0700, Grant Kelly wrote: It's an immutable page. Can someone who already has access make the edit? Umm, are you sure? I don't see any ACL's on the page. (Though you need to register an account on the wiki before editing.) -- Pauli Virtanen ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] Home for pyhdf5io?
Dag Sverre Seljebotn wrote: albert.thuswald...@gmail.com wrote: Dear list, I'm writing this because i have developed a small python module that might be of interest to you the readers of this list: http://code.google.com/p/pyhdf5io/ It basically implements load/save functions that mimic the behaviour of those found in Matlab, i.e. with them you can store your variables from within the interactive shell (IPython, python) or from within a function, and then load them back in again. One important difference is that the hdf5 format is used to store the variables, which comes with a a number of benefits: - a open standard file format which is supported by many applications. - completely portable file format across different platforms. Read more here: http://www.hdfgroup.org/HDF5/whatishdf5.html And now to the question: I think that this module is to small to be developed and maintained on its on, I think It would be better if it could be part of some larger project. So where would pyhdf5io fit in? Any tips and ideas are highly appreciate I'd expect to find it in http://h5py.alfven.org/ I think... Please disregard this, I didn't notice you had a PyTable dependency. -- Dag Sverre ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] Home for pyhdf5io?
Hi Albert, So this is a wrapper on top of PyTables to implement load() and save()? Neat. Obviously if you're installing PyTables, you can do a lot better and organize your data hierarchically without the messiness of Matlab structures, walk the node tree, all kinds of fun stuff, but if you're an expatriate matlab user and just want to save some matrices... this is great. Notably, that was one of my gripes about ipython+numpy+scipy +matplotlib when I first came from Matlab. I think you should send a message to the PyTables list, ask Francesc if he thinks it has a place in PyTables for it as a 'lite' wrapper or something, for people who need to save data but don't need/are intimidated by all the features that PyTables provides. David On 21-May-09, at 4:04 PM, albert.thuswald...@gmail.com wrote: Dear list, I'm writing this because i have developed a small python module that might be of interest to you the readers of this list: http://code.google.com/p/pyhdf5io/ It basically implements load/save functions that mimic the behaviour of those found in Matlab, ie with them you can store your variables from within the interactive shell (IPython, python) or from within a function, and then load them back in again. One important difference is that the hdf5 format is used to store the variables, which comes with aa number of benefits: - a open standard file format which is supported by many applications. - completely portable file format across different platforms. Read more here: http://www.hdfgroup.org/HDF5/whatishdf5.html And now to the question: I think that this module is to small to be developed and maintained on its on, I think It would be better if it could be part of some larger project. So where would pyhdf5io fit in? Any tips and ideas are highly appreciated. Thanks. /Albert ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] Home for pyhdf5io?
On Thu, May 21, 2009 at 22:38, David Warde-Farley d...@cs.toronto.edu wrote: Hi Albert, So this is a wrapper on top of PyTables to implement load() and save()? Neat. Yes, you got the idea. in its most simplest form you can type: hdf5save() And all your local variables are saved to a file with the default file name hdf5io.h5. Of course it also allows you to specify a file name and what variables you would like to save. As it is based on hdf5 you can also store the variables to a certain group within the file (If you know how hdf5 works, you probably know what I'm talking about). Appending data to existing hdf5-files is also possible. Obviously if you're installing PyTables, you can do a lot better and organize your data hierarchically without the messiness of Matlab structures, walk the node tree, all kinds of fun stuff, but if you're an expatriate matlab user and just want to save some matrices... this is great. Notably, that was one of my gripes about ipython+numpy+scipy +matplotlib when I first came from Matlab. Exactly! I think you should send a message to the PyTables list, ask Francesc if he thinks it has a place in PyTables for it as a 'lite' wrapper or something, for people who need to save data but don't need/are intimidated by all the features that PyTables provides. Actually, I just e-mail Francesc, see what he thinks. Thanks for your reply. Also thanks to the others who also have replied /Albert David On 21-May-09, at 4:04 PM, albert.thuswald...@gmail.com wrote: Dear list, I'm writing this because i have developed a small python module that might be of interest to you the readers of this list: http://code.google.com/p/pyhdf5io/ It basically implements load/save functions that mimic the behaviour of those found in Matlab, ie with them you can store your variables from within the interactive shell (IPython, python) or from within a function, and then load them back in again. One important difference is that the hdf5 format is used to store the variables, which comes with aa number of benefits: - a open standard file format which is supported by many applications. - completely portable file format across different platforms. Read more here: http://www.hdfgroup.org/HDF5/whatishdf5.html And now to the question: I think that this module is to small to be developed and maintained on its on, I think It would be better if it could be part of some larger project. So where would pyhdf5io fit in? Any tips and ideas are highly appreciated. Thanks. /Albert ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion