Mailman 3 August 2004 - NumPy-Discussion

fromstring doesn't accept buffer object
by Nadav Horesh Aug. 26, 2004

Aug. 26, 2004

Numerc accepts a buffer object instead of a string: arr = Numeric.fromstring(buffer(.some_string), ..) while numarray doesn't. Is it intentional? Nadav.

2 1

Accessing a C library from numarray
by Bruce Southey Aug. 25, 2004

Aug. 25, 2004

Hi, I need to access a GLP'ed C library in numarray. The library functions are scalar ints and doubles as inputs and usually doubles as outputs. Is there a recommended (and simple) way to achieve this? There are a few different C API's that exist from the Python to numarrays. Would it be sufficient to first write C code that uses the library and use that code for a C API in Python or numarray? Thanks, Bruce Southey

2 1

MA.zeros problem
by smpitts＠ou.edu Aug. 23, 2004

Aug. 23, 2004

Hi all, I have some code that uses Numeric, and I'm trying to change it so that it supports MA as well. My code uses Numeric.zeros to create a matrix and then populates it, but I'm having trouble figuring out how to do something similar in MA. Python 2.2.3 (#1, Oct 15 2003, 23:33:35) [GCC 3.3.1 20030930 (Red Hat Linux 3.3.1-6)] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> import MA >>> MA.__version__ '11.1.0' >>> a = MA.zeros((2,2)) >>> a array( [[0,0,] [0,0,]]) >>> a[1][1] = 2 >>> a array( [[0,0,] [0,0,]]) Is there some function to automatically create a masked array of arbitrary dimensions filled with zeroes? Thanks. -- Stephen Pitts smpitts(a)ou.edu

2 1

Re: [Numpy-discussion] Empty arrays strange behaviour
by Todd Miller Aug. 18, 2004

Aug. 18, 2004

On Wed, 2004-08-18 at 11:42, Tim Hochberg wrote: > Todd Miller wrote: > > > [CHOP] > > >The key points for me are: > > > >1. The slice operation doesn't raise an exception even though it has > >"invalid" indexes. As Tim H. pointed out, those are standard Python > >semantics. > > > >2. The shape of the resultant array is the same. > > > >I don't know anything about the purpose of "zeros()" in this context. It > >looks inconsistent to me. > > > > > > > I believe it's to disambiguate the various possible arrays of total size > zero. I like the Numeric behaviour better here, although it's not a big > deal. In the following, both the size and type of the numarrays are > ambiguous from their repr(). This means that among other things > eval(repr(arg)) won't return arg even for nice short arrays (I realize > the numarray truncates the reprs of long arrays by default and I think > that's fine -- I imagine there's some way to disable it, but I haven't > looked for it yet). > > -tim > > >>> import Numeric as np > >>> np.zeros([0]) > zeros((0,), 'l') > >>> np.zeros([0,2]) > zeros((0, 2), 'l') > >>> import numarray as na > >>> na.zeros([0]) > array([]) > >>> na.zeros([0,2]) > array([]) > >>> na.arange(100) > > OK, I logged this on SF. Thanks for the explanation. Todd

1 0

RE: [Numpy-discussion] Empty arrays strange behaviour
by Nadav Horesh Aug. 18, 2004

Aug. 18, 2004

-----Original Message----- From: Todd Miller [mailto:jmiller@stsci.edu] Sent: Wed 18-Aug-04 13:46 To: Nadav Horesh Cc: numpy-discussion Subject: Re: [Numpy-discussion] Empty arrays strange behaviour On Wed, 2004-08-18 at 07:36, Nadav Horesh wrote: > >>> a = arange(20, shape=(4,5)) > >>> a > array([[ 0, 1, 2, 3, 4], > [ 5, 6, 7, 8, 9], > [10, 11, 12, 13, 14], > [15, 16, 17, 18, 19]]) > >>> a[2:4,5:7] # Shouldn't it raise index error? > array([]) I wasn't sure about this, but what we're doing is faithful to what Numeric does. > >>> a[2:4, 5:7].iscontiguous() > 0 # ???? Likewise here. > >>> a[2:4, 5:7].copy().iscontiguous() > 0 # ???? We're not compatible here. > Shouldn't iscontiguous method of empty arrays and array.copy() return 1? I'd say we should go with compatibility. Todd It isn't a major flaw, so compatibility is OK. I got to this issue since I had a bug. If there was an option to get an IndexError exception it would ease the debugging. Thank you, Nadav.

4 3

Bool array missing sum function.
by Nadav Horesh Aug. 18, 2004

Aug. 18, 2004

Why Bool arrays are missing the sum method? >>> b = array((1,0,1), type=Bool) >>> b array([1, 0, 1], type=Bool) >>> b.sum() Traceback (most recent call last): File "<pyshell#129>", line 1, in -toplevel- b.sum() File "/usr/local/lib/python2.4/site-packages/numarray/numarraycore.py", line 1125, in sum return ufunc.add.reduce(ufunc.add.areduce(self, type=type).flat, type=type) File "/usr/local/lib/python2.4/site-packages/numarray/ufunc.py", line 1165, in _cum_cache_miss mode, win1, wout, cfunc, ufargs = \ File "/usr/local/lib/python2.4/site-packages/numarray/ufunc.py", line 1225, in _cum_setup otype, cfunc = ufdict[(optype,)] KeyError: Bool Nadav.

2 1

Empty arrays strange behaviour
by Nadav Horesh Aug. 17, 2004

Aug. 17, 2004

>>> a = arange(20, shape=(4,5)) >>> a array([[ 0, 1, 2, 3, 4], [ 5, 6, 7, 8, 9], [10, 11, 12, 13, 14], [15, 16, 17, 18, 19]]) >>> a[2:4,5:7] # Shouldn't it raise index error? array([]) >>> a[2:4, 5:7].iscontiguous() 0 # ???? >>> a[2:4, 5:7].copy().iscontiguous() 0 # ???? Shouldn't iscontiguous method of empty arrays and array.copy() return 1? Nadav.

2 1

core dump during import
by Skip Montanaro Aug. 17, 2004

Aug. 17, 2004

(I posted this last week to scipy-user but received no responses. Since it seems to involve Numeric I thought I'd try here.) ------------------------------------------------------------------------------ I'm the "python guy" at work which by default makes me the "scipy installation guy" as well, though I'm not using it for any of my own projects. Awhile ago I performed a very basic installation: install latest f2py install Numeric 23.1 install SciPy_complete-0.3 The platform is Solaris 9 on Intel. I'm using Python 2.3.4 (though originally installed using 2.3.3). GCC and G77 are both at 3.3.2. The installation goes fine with the one caveat that I needed to point LD_RUN_PATH at the g77 lib dir so libg2c.so.0 could be found at runtime. If I execute from scipy.stats import distribution at the interpreter prompt I get a segfault. Here's the beginning of the gdb traceback: #0 0x20ec8356 in ?? () #1 0xdd2d7e24 in Cephes_InitOperators (dictionary=0x8486dfc) at Lib/special/cephesmodule.c:413 #2 0xdd2d7a7a in initcephes () at Lib/special/cephesmodule.c:1105 #3 0x080d082e in _PyImport_LoadDynamicModule ( name=0x8040500 "scipy.special.cephes", pathname=0x8040070 "/opt/lang/python/lib/python2.3/site-packages/scipy/special/cephes.so", fp=0x8148780) at ../Python/importdl.c:53 #4 0x080cd637 in load_module (name=0x8040500 "scipy.special.cephes", fp=0x8148780, buf=0x8040070 "/opt/lang/python/lib/python2.3/site-packages/scipy/special/cephes.so", type=3, loader=0x0) at ../Python/import.c:1708 #5 0x080ce345 in import_submodule (mod=0x8149fa4, subname=0x804050e "cephes", fullname=0x8040500 "scipy.special.cephes") at ../Python/import.c:2291 #6 0x080cdea4 in load_next (mod=0x8149fa4, altmod=0x81265e0, p_name=0x8040930, buf=0x8040500 "scipy.special.cephes", p_buflen=0x80404fc) at ../Python/import.c:2111 #7 0x080cfbf1 in import_module_ex (name=0x0, globals=0x8486c64, locals=0x8486c64, fromlist=0x83e256c) at ../Python/import.c:1957 #8 0x080ced40 in PyImport_ImportModuleEx (name=0x8369c34 "cephes", globals=0x8486c64, locals=0x8486c64, fromlist=0x83e256c) at ../Python/import.c:1998 #9 0x080a1bde in builtin___import__ (self=0x0, args=0x8483d4c) at ../Python/bltinmodule.c:45 #10 0x080fac78 in PyCFunction_Call (func=0x8151bec, arg=0x8483d4c, kw=0x0) at ../Objects/methodobject.c:108 #11 0x08062e90 in PyObject_Call (func=0x8151bec, arg=0x8483d4c, kw=0x0) at ../Objects/abstract.c:1755 Since I don't use SciPy, Cephes or f2py outside of installing SciPy I don't really know where to start poking. Frame 1 is at this line: 413 f = PyUFunc_FromFuncAndData(cephes3a_functions, bdtr_data, cephes_4_types, 2, 3, 1, PyUFunc_None, "bdtr", bdtr_doc, 0); and cephes3a_functions looks suspicious: (gdb) p cephes3a_functions $2 = {0x5, 0x6} I presume considering the variable name those are supposed to be valid function pointers. Stepping through initcephes leads me to this: (gdb) n 393 cephes3a_functions[0] = PyUFunc_fff_f_As_iid_d; (gdb) n 391 cephes3_functions[2] = PyUFunc_ffF_F_As_ddD_D; (gdb) n 392 cephes3_functions[3] = PyUFunc_ddD_D; (gdb) n 394 cephes3a_functions[1] = PyUFunc_ddd_d_As_iid_d; (gdb) n 395 cephes3_2_functions[0] = PyUFunc_fff_ff_As_ddd_dd; (gdb) p cephes3a_functions $9 = {0x5, 0x6} (gdb) p PyUFunc_ddD_D $10 = {void (char **, int *, int *, void *)} 0xdd2e4720 <PyUFunc_ddD_D> (gdb) p PyUFunc_ddd_d_As_iid_d $11 = {void (char **, int *, int *, void *)} 0xdd2e4a20 <PyUFunc_ddd_d_As_iid_d> It seems there's something amiss with the cephes3a_functions initialization. In fact, other cephes*_functions variables appear bogus as well: (gdb) p cephes3_functions $12 = {0, 0, 0x3, 0x4} (gdb) p cephes1_functions $13 = {0x464c457f, 0x10101} (gdb) p cephes2_4_functions $14 = {0x7e34, 0x7} Does this ring a bell with anybody? Any clues about how to proceed would be appreciated. Thanks, -- Skip Montanaro Got spam? http://www.spambayes.org/ skip(a)pobox.com

1 0

Using sum() within a function
by Jim Cser Aug. 17, 2004

Aug. 17, 2004

Hello- I have a function to generate a multi-dimensional array, which then gets summed over one axis. The problem is that the dimensions are large, and I run out of memory when I create the entire array, so I'm trying to do the sum *within* the function. Example-- variables x,y,z,t; dimensions numX, numY, numZ, numT; functions f1(x,y,z,t), f2(y,z,t); want to calculate f1*f2 and sum over t to get out[x,y,z]. With loops, I could do it like-- out = zeros((numX,numY,numZ)) for x in range(numX): for y in range(numY): for z in range(numZ): for t in range(numT): tempval = f1(x,y,z,t) * f2(y,z,t) out[x,y,z] = out[x,y,z] + tempval With numarray, if I had enough memory, I could just do-- temp1 = fromfunction(f1,(numX,numY,numZ,numT)) temp2 = resize(fromfunction(f2,(numY,numZ,numT)),(numX,numY,numZ,numT)) out = sum(temp1 * temp2, axis = 3) Instead, I'm trying to do something like-- def f3(x,y,z): for t in range(numT): tempval = f1(x,y,z,t) * f2(y,z,t) outval = sum(tempval,axis = 3) return outval out = fromfunction(f3,(numX,numY,numZ)) I've been trying various slicing and indexing, but I can't seem to get that *extra* dimension within the 3-D function. I've scoured the documentation and list archives, but haven't found exactly what I need. Any suggestions? Am I stuck with generating the entire 4-D array? Thanks in advance, Jim Cser

3 5

Re: core dump during import
by Skip Montanaro Aug. 17, 2004

Aug. 17, 2004

A few days ago on scipy-user and earlier today on numpy-discussion I wrote: If I execute from scipy.stats import distribution at the interpreter prompt I get a segfault. Stepping through the suspect code yielded very strange results, so I tried recompiling cephesmodule.c with -O0. That seems to have solved the problem for now. I guess that means it's a GCC optimizer bug. -- Skip Montanaro Got spam? http://www.spambayes.org/ skip(a)pobox.com

1 0