Mailman 3 September 2010 - NumPy-Discussion

common significant diigits
by Benjamin Root 16 Sep '10

16 Sep '10

Hello, I am trying to solve a problem in matplotlib where I would have an array of floating point numbers and I want to quickly determine what is the closest common offset to a power of 10. In other words, if given: [12373.43, 12375.89, 12370.18], I would want returned something like either 12370.0, or the lowest common order of magnitude (in this case, 10). Is there some sort of neat math/numpy trick to figure this out? I already have a brute-force method with a while loop, but I am looking for something a little bit more elegant. Thanks, Ben Root

3 3

array view with different shape
by Mark Bakker 16 Sep '10

16 Sep '10

Hello List, Can I make a view of an entire array but with a different shape? For example: a = zeros((2,3,4)) Now I want array b to be identical to a, but with shape (2,12). b = a; b.shape = (2,12) This doesn't work, of course, as also the shape of a changes. I know I can simply make a copy of a and change the shape, but can I make b to be a view of a (so that when I change a, then b changes as well)? Thanks, Mark

3 2

Re: [Numpy-discussion] common significant diigits
by Mark Bakker 16 Sep '10

16 Sep '10

10*np.floor(a/10).min()

1 0

Re: [Numpy-discussion] ANN: OpenOpt 0.31, FuncDesigner 0.21, DerApproximator 0.21
by Mark Bakker 16 Sep '10

16 Sep '10

Dmitrey, Great to hear OpenOpt development is moving forward. I tried to easy_install on both a Mac and on Windows, but both installations crashed. With similar (maybe identical) errors on "atexit" commands. Any solution? Mark

1 0

Python2.4 and date-time - tickets 1475 and 1613
by Bruce Southey 15 Sep '10

15 Sep '10

Hi, Datetime support in Numpy requires ctypes - RuntimeError is raised because ctypes has only been included in Python starting with Python2.5. This Python2.4 dependency must be included in the documentation so I created ticket 1613 (http://projects.scipy.org/numpy/ticket/1613). A standalone version of ctypes is available for Python2.3 and Python2.4: http://starship.python.net/crew/theller/ctypes/ The numpy tests pass after installing ctypes under Python2.4. I think ticket 1475 is fixed by skipping that test in a similar way to the skipped tests in 'numpy/tests/test_ctypeslib.py'. Bruce PS A big thanks for fixing the other Python2.4 errors I mentioned with regards to scipy!

1 0

ANN: OpenOpt 0.31, FuncDesigner 0.21, DerApproximator 0.21
by Dmitrey 15 Sep '10

15 Sep '10

Hi all, I'm glad to inform you about new releases: OpenOpt 0.31, FuncDesigner 0.21, DerApproximator 0.21 For details see http://forum.openopt.org/viewtopic.php?id=299 or visit our homepage http://openopt.org Regards, Dmitrey

1 0

test for long double support
by Neal Becker 15 Sep '10

15 Sep '10

What is the recommended way to test for long double and clongdouble? Is it: #if NPY_SIZEOF_LONGDOUBLE == NPY_SIZEOF_DOUBLE

3 3

Buildbot errors and failures
by Charles R Harris 15 Sep '10

15 Sep '10

Just thought I'd throw these out there, they've been showing for a while. Windows XP: =========== ERROR: test_fromfile_tofile_seeks (test_regression.TestRegression) ---------------------------------------------------------------------- Traceback (most recent call last): File "C:\buildbot\numpy\b11\numpy-install25\Lib\site-packages\numpy\core\tests\test_regression.py", line 1384, in test_fromfile_tofile_seeks ret = np.fromfile(f, count=4, dtype='u1') IOError: first argument must be an open file BSD : ===== gmake test /usr/local/bin/python2.6 -c 'import numpy,sys;sys.exit(not numpy.test("full", verbose=10).wasSuccessful())' Traceback (most recent call last): File "<string>", line 1, in <module> File "/home/doconnor/numpy-buildbot/b12/numpy-install/lib/python2.6/site-packages/numpy/__init__.py", line 136, in <module> import add_newdocs File "/home/doconnor/numpy-buildbot/b12/numpy-install/lib/python2.6/site-packages/numpy/add_newdocs.py", line 9, in <module> from numpy.lib import add_newdoc File "/home/doconnor/numpy-buildbot/b12/numpy-install/lib/python2.6/site-packages/numpy/lib/__init__.py", line 13, in <module> from polynomial import * File "/home/doconnor/numpy-buildbot/b12/numpy-install/lib/python2.6/site-packages/numpy/lib/polynomial.py", line 17, in <module> from numpy.linalg import eigvals, lstsq File "/home/doconnor/numpy-buildbot/b12/numpy-install/lib/python2.6/site-packages/numpy/linalg/__init__.py", line 48, in <module> from linalg import * File "/home/doconnor/numpy-buildbot/b12/numpy-install/lib/python2.6/site-packages/numpy/linalg/linalg.py", line 23, in <module> from numpy.linalg import lapack_lite ImportError: /usr/local/lib/liblapack.so.4: Undefined symbol "_gfortran_st_write" Chuck

1 0

sf.net export controls for numpy/scipy
by Jarrod Millman 15 Sep '10

15 Sep '10

Hello, I plan to update the export controls settings for both numpy and scipy to: "This project does NOT incorporate, access, call upon, or otherwise use encryption of any kind, including, but not limited to, open source algorithms and/or calls to encryption in the operating system or underlying platform." Unless I hear from someone that I overlooked something, I will make the changes tomorrow. Best, Jarrod

1 0

[OT: MATLAB] Any way to globally make Matlab struct attributes Python-property-like
by David Goldsmith 15 Sep '10

15 Sep '10

I.e., I'd, at minimum, like to globally replace get(Handel, 'Property') with object.Property and set(Handel, 'Property', value) with object.Property = value to an arbitrary level of composition. (It's really getting cumbersome having to compound gets and sets all over the place while debugging.) (Sorry for the OT post; I thought I'd get a more sympathetic response here than on the MATLAB lists.) ;) Thanks! DG -- In science it often happens that scientists say, 'You know that's a really good argument; my position is mistaken,' and then they would actually change their minds and you never hear that old view from them again. They really do it. It doesn't happen as often as it should, because scientists are human and change is sometimes painful. But it happens every day. I cannot recall the last time something like that happened in politics or religion. - Carl Sagan, 1987 CSICOP address

5 5