Hi,<br><br>I'm pleased to announce the availability of the first release candidate of NumPy 1.6.2. This is a maintenance release. Due to the delay of the NumPy 1.7.0, this release contains far more fixes than a regular NumPy bugfix release. It also includes a number of documentation and build improvements.<br>
<br>Sources and binary installers can be found at <a href="https://sourceforge.net/projects/numpy/files/NumPy/1.6.2rc1/">https://sourceforge.net/projects/numpy/files/NumPy/1.6.2rc1/</a><br><br>Please test this release and report any issues on the numpy-discussion mailing list.<br>
<br>Cheers,<br>Ralf<br><br><br><br>``numpy.core`` issues fixed<br>---------------------------<br><br>#2063 make unique() return consistent index<br>#1138 allow creating arrays from empty buffers or empty slices<br>#1446 correct note about correspondence vstack and concatenate<br>
#1149 make argmin() work for datetime<br>#1672 fix allclose() to work for scalar inf<br>#1747 make np.median() work for 0-D arrays<br>#1776 make complex division by zero to yield inf properly<br>#1675 add scalar support for the format() function<br>
#1905 explicitly check for NaNs in allclose()<br>#1952 allow floating ddof in std() and var()<br>#1948 fix regression for indexing chararrays with empty list<br>#2017 fix type hashing<br>#2046 deleting array attributes causes segfault<br>
#2033 a**2.0 has incorrect type<br>#2045 make attribute/iterator_element deletions not segfault<br>#2021 fix segfault in searchsorted()<br>#2073 fix float16 __array_interface__ bug<br><br><br>``numpy.lib`` issues fixed<br>
--------------------------<br><br>#2048 break reference cycle in NpzFile<br>#1573 savetxt() now handles complex arrays<br>#1387 allow bincount() to accept empty arrays<br>#1899 fixed histogramdd() bug with empty inputs<br>
#1793 fix failing npyio test under py3k<br>#1936 fix extra nesting for subarray dtypes<br>#1848 make tril/triu return the same dtype as the original array<br>#1918 use Py_TYPE to access ob_type, so it works also on Py3<br>
<br><br>``numpy.f2py`` changes<br>----------------------<br><br>ENH: Introduce new options extra_f77_compiler_args and extra_f90_compiler_args<br>BLD: Improve reporting of fcompiler value<br>BUG: Fix f2py test_kind.py test<br>
<br><br>``numpy.poly`` changes<br>----------------------<br><br>ENH: Add some tests for polynomial printing<br>ENH: Add companion matrix functions <br>DOC: Rearrange the polynomial documents<br>BUG: Fix up links to classes<br>
DOC: Add version added to some of the polynomial package modules <br>DOC: Document xxxfit functions in the polynomial package modules <br>BUG: The polynomial convenience classes let different types interact<br>
DOC: Document the use of the polynomial convenience classes<br>DOC: Improve numpy reference documentation of polynomial classes<br>ENH: Improve the computation of polynomials from roots <br>STY: Code cleanup in polynomial [*]fromroots functions <br>
DOC: Remove references to cast and NA, which were added in 1.7<br><br><br>``numpy.distutils`` issues fixed<br>-------------------------------<br><br>#1261 change compile flag on AIX from -O5 to -O3<br>#1377 update HP compiler flags<br>
#1383 provide better support for C++ code on HPUX<br>#1857 fix build for py3k + pip<br>BLD: raise a clearer warning in case of building without cleaning up first<br>BLD: follow build_ext coding convention in build_clib<br>
BLD: fix up detection of Intel CPU on OS X in system_info.py<br>BLD: add support for the new X11 directory structure on Ubuntu & co.<br>BLD: add ufsparse to the libraries search path. <br>BLD: add 'pgfortran' as a valid compiler in the Portland Group <br>
BLD: update version match regexp for IBM AIX Fortran compilers.<br><br>