Mailman 3 March 2010 - NumPy-Discussion

basic numpy array operation
by gerardo.berbeglia 15 Mar '10

15 Mar '10

Suppose i have an array A of length n I want a variable b to be an array consisting of the first k elements of A and variable c to be an array with the last n-k elements of A. How do you do this? Example A = np.array([1,2,3,4,5,6]), k = 2 b = [1,2] c=[3,4,5,6] Thanks. -- View this message in context: http://old.nabble.com/basic-numpy-array-operation-tp27908290p27908290.html Sent from the Numpy-discussion mailing list archive at Nabble.com.

2 1

Numpi array and typemap
by Michel Dupront 15 Mar '10

15 Mar '10

Hello, I am trying to do what was suppose to be a very easy exercice. It consists in converting a C++ class Container into a numpy array as show below using typename My problem is that if I make the method getContainer run from python I will get a numpy array as expected by with wrong values. I guess I have a memory problem but I do not understand exactly wwhat is going on. I was hoping that somebody could help me with that and maybe tell me how to use the PyArray_SimpleNewFromData macro. Thanks %module Foo %{ #define SWIG_FILE_WITH_INIT #include "container.h" template <class T> Container<T> getContainer() { ... C Container construction that will have a pointer to (1,2,3,4,5)... return C; } %} %typedef unsigned int size_t; %include "numpy.i" %init %{ import_array(); %} template <class T> class Container { public: Container(size_t length, T* pfirst_value); size_t lentgh() { ... return the lentgh ...} T* ptr() { ... return pointer to the first value ... } }; %template(cN) Container<size_t>; %typemap(python,out) Container<size_t> { npy_intp dims[] = { 5 }; PyObject* obj = PyArray_SimpleNewFromData (1,dims,PyArray_UINT,$1.ptr()); $result = obj; } template <class T> navlib::array<T> getContainer(); %template(getContainerN ) getContainer<size_t>; _________________________________________________________________ Découvrez comment SURFER DISCRETEMENT sur un site de rencontres ! http://clk.atdmt.com/FRM/go/206608211/direct/01/

2 1

long integers in genfromtxt
by josef.pktd＠gmail.com 15 Mar '10

15 Mar '10

I was trying to find out what the "helpful" message "TypeError: expected a readable buffer object" means and it seems genfromtxt has problems identifying long integers (at least on Windows 32) >>> np.array(4160680000,int) Traceback (most recent call last): File "<pyshell#4>", line 1, in <module> np.array(4160680000,int) OverflowError: long int too large to convert to int >>> s = '''Date,Open,High,Low,Close,Volume,Adj Close ... 2010-02-12,1075.95,1077.81,1062.97,1075.51,4160680000,1075.51 ... 2010-02-11,1067.10,1080.04,1060.59,1078.47,4400870000,1078.47''' >>> sh = StringIO(s) >>> data = np.genfromtxt(sh, delimiter=",", dtype=None, names=True) Traceback (most recent call last): File "C:\Programs\Python25\Lib\site-packages\numpy\lib\io.py", line 1367, in genfromtxt output = np.array(data, dtype=ddtype) TypeError: expected a readable buffer object >>> sh = StringIO(s) >>> data = ts.tsfromtxt(sh, delimiter=",", dtype=None, datecols=(0,), names=True) Traceback (most recent call last): File "\Programs\Python25\Lib\site-packages\scikits\timeseries\extras.py", line 504, in tsfromtxt File "\Programs\Python25\Lib\site-packages\scikits\timeseries\_preview.py", line 1312, in genfromtxt TypeError: expected a readable buffer object >>> dt= [('','S10'),('',float),('',float),('',float),('',float),('',int),('',float)] >>> sh = StringIO(s) >>> data = np.genfromtxt(sh, delimiter=",", dtype=dt, names=True) Traceback (most recent call last): File "C:\Programs\Python25\Lib\site-packages\numpy\lib\io.py", line 1388, in genfromtxt rows = np.array(data, dtype=[('', _) for _ in dtype_flat]) TypeError: expected a readable buffer object >>> sh = StringIO(s) >>> data = ts.tsfromtxt(sh, delimiter=",", dtype=dt, datecols=(0,), names=True) Traceback (most recent call last): File "\Programs\Python25\Lib\site-packages\scikits\timeseries\extras.py", line 451, in tsfromtxt File "\Programs\Python25\Lib\site-packages\scikits\timeseries\_preview.py", line 798, in easy_dtype File "\Programs\Python25\Lib\site-packages\scikits\timeseries\_preview.py", line 364, in __call__ File "\Programs\Python25\Lib\site-packages\scikits\timeseries\_preview.py", line 329, in validate TypeError: object of type 'bool' has no len() using float works, but I needed the debugger to figure out what the problem with my data is (or whether I just make mistakes) >>> dt= [('','S10'),('',float),('',float),('',float),('',float),('',float),('',float)] >>> sh = StringIO(s) >>> data = np.genfromtxt(sh, delimiter=",", dtype=dt, names=True) >>> data array([ ('2010-02-12', 1075.95, 1077.8099999999999, 1062.97, 1075.51, 4160680000.0, 1075.51), ('2010-02-11', 1067.0999999999999, 1080.04, 1060.5899999999999, 1078.47, 4400870000.0, 1078.47)], dtype=[('Date', '|S10'), ('Open', '<f8'), ('High', '<f8'), ('Low', '<f8'), ('Close', '<f8'), ('Volume', '<f8'), ('Adj_Close', '<f8')]) >>> >>> ts.version.version '0.91.3' Josef

3 3

matrix operation.
by gerardo.berbeglia 15 Mar '10

15 Mar '10

I have another matrix operations which seems a little more complicated. Let A be an n x n matrix and let S be a subset of {0,...,n-1}. Assume S is represented by a binary vector s, with a 1 at the index i if i is in S. (e.g. if S={0,3} then s = [1,0,0,1]) I re-post the question because the example was wrong. I would like to have an efficient way to compute the function B = f (A,S) characterized as follows: - For each column i such that i is in S, then the column i of B is equal to the column i of A. - For each column i such that i is NOT in S, then the column i of B is equal to the ith column of the n x n identity matrix. Example. n=4. A = [[2,3,4,5], [3,4,5,6] [4,5,6,7] [5,6,7,8]] S = {0,2} => s=[1,0,1,0] f(A,S) = [[2,0,4,0], [3,1,5,0], [4,0,6,0], [5,0,7,1]] Which is the best way to compute f? Thanks again. -- View this message in context: http://old.nabble.com/matrix-operation.-tp27905661p27905661.html Sent from the Numpy-discussion mailing list archive at Nabble.com.

1 0

more complex matrix operation
by Gerardo Berbeglia 15 Mar '10

15 Mar '10

I have another matrix operations which seems a little more complicated. Let A be an n x n matrix and let S be a subset of {0,...,n-1}. Assume S is represented by a binary vector s, with a 1 at the index i if i is in S. (e.g. if S={0,3} then s = [1,0,0,1]) I would like to have an efficient way to compute the function B = f (A,S) characterized as follows: - For each column i such that i is in S, then the column i of B is equal to the column i of A. - For each column i such that i is NOT in S, then the column i of B is equal to the ith column of the n x n identity matrix. Example. n=4. A = [[2,2,2,2],[3,3,3,3][4,4,4,4][5,5,5,5]] S = {0,2} => s=[1,0,1,0] f(A,S) = [[2,2,2,2],[0,1,0,0],[4,4,4,4],[0,0,0,1]] Which is the best way to compute f? Thanks again.

3 2

another matrix operation
by Gerardo Berbeglia 15 Mar '10

15 Mar '10

I would like to do with numpy the following operation. Let A be an n x n matrix and let s be an integer between 1 and n. I would like to have an n x n matrix B = f(A,s) such that - If we only look at the first s columns of B, we will not see any difference with respect to the first s columns of the n x n identity matrix. - For the last n-s columns of B, we will not see any difference with respect to the last n-s columns of A. Example. n = 4, s = 2 A= [[2,2,2,2],[3,3,3,3],[4,4,4,4],[5,5,5,5]] B= f(A,s) = [[1,0,2,2],[0,1,3,3],[0,0,4,4],[0,0,5,5]] Is there a way to do this efficiently (without loops)? Thanks in advance.

2 2

matrix division without a loop
by gerardo.berbeglia 12 Mar '10

12 Mar '10

Hello, I want to "divide" an n x n (2-dimension) numpy array matrix A by a n (1-dimension) array d as follows: Take n = 2. Let A= 2 3 1 10 and let d = [ 3 2 ] Then i would like to have "A/d" = 2/3 3/3 1/2 10/2 This is to avoid loops to improve the speed. Thank you in advance! -- View this message in context: http://old.nabble.com/matrix-division-without-a-loop-tp27881350p27881350.ht… Sent from the Numpy-discussion mailing list archive at Nabble.com.

3 2

matrix operation
by gerardo.berbeglia 12 Mar '10

12 Mar '10

Hello, I want to "divide" an n x n (2-dimension) numpy array matrix A by a n (1-dimension) array d as follows: Take n = 2. Let A= 2 3 1 10 and let d = [ 3 2 ] Then i would like to have "A/d" = 2/3 3/3 1/2 10/2 This is to avoid loops to improve the speed. Thank you in advance! -- View this message in context: http://old.nabble.com/matrix-operation-tp27879913p27879913.html Sent from the Numpy-discussion mailing list archive at Nabble.com.

2 1

crash at prompt exit after running test
by Johann Cohen-Tanugi 12 Mar '10

12 Mar '10

hi there, I just installed the current head of numpy and built it. trying python and then import numpy, and then CTRL-D to exit, all goes well. But doing the same with a numpy.test() before CTRL-D ends up in : <clip> Ran 2892 tests in 35.814s OK (KNOWNFAIL=4, SKIP=6) <nose.result.TextTestResult run=2892 errors=0 failures=0> >>> python: Modules/gcmodule.c:277: visit_decref: Assertion `gc->gc.gc_refs != 0' failed. Aborted (core dumped) Does that ring a bell to any of you? My system is Fedora 12, I used gfortran to build and In [10]: sys.path Out[10]: ['', '/home/cohen/.local/bin', '/home/cohen/sources/python/ipython', '/usr/lib/python26.zip', '/usr/lib/python2.6', '/usr/lib/python2.6/plat-linux2', '/usr/lib/python2.6/lib-tk', '/usr/lib/python2.6/lib-old', '/usr/lib/python2.6/lib-dynload', '/home/cohen/.local/lib/python2.6/site-packages', '/usr/lib/python2.6/site-packages', '/usr/lib/python2.6/site-packages/PIL', '/usr/lib/python2.6/site-packages/gst-0.10', '/usr/lib/python2.6/site-packages/gtk-2.0', '/usr/lib/python2.6/site-packages/webkit-1.0', '/usr/lib/python2.6/site-packages/wx-2.8-gtk2-unicode', '/home/cohen/sources/python/ipython/IPython/extensions', u'/home/cohen/.ipython'] numpy and ipython live in $HOME/.local thanks in advance, Johann

6 24

Re: [Numpy-discussion] arange including stop value?
by Tom K. 12 Mar '10

12 Mar '10

davefallest wrote: > > ... > In [3]: np.arange(1.01, 1.1, 0.01) > Out[3]: array([ 1.01, 1.02, 1.03, 1.04, 1.05, 1.06, 1.07, 1.08, > 1.09, 1.1 ]) > > Why does the ... np.arange command end up including my stop value? > >From the help for arange: For floating point arguments, the length of the result is ``ceil((stop - start)/step)``. Because of floating point overflow, this rule may result in the last element of `out` being greater than `stop`. -- View this message in context: http://old.nabble.com/arange-including-stop-value--tp27866607p27872069.html Sent from the Numpy-discussion mailing list archive at Nabble.com.

2 1