|
| 1 | +========================= |
| 2 | +NumPy 1.6.2 Release Notes |
| 3 | +========================= |
| 4 | + |
| 5 | +This is a bugfix release in the 1.6.x series. Due to the delay of the NumPy |
| 6 | +1.7.0 release, this release contains far more fixes than a regular NumPy bugfix |
| 7 | +release. It also includes a number of documentation and build improvements. |
| 8 | + |
| 9 | + |
| 10 | +``numpy.core`` issues fixed |
| 11 | +--------------------------- |
| 12 | + |
| 13 | +#2063 make unique() return consistent index |
| 14 | +#1138 allow creating arrays from empty buffers or empty slices |
| 15 | +#1446 correct note about correspondence vstack and concatenate |
| 16 | +#1149 make argmin() work for datetime |
| 17 | +#1672 fix allclose() to work for scalar inf |
| 18 | +#1747 make np.median() work for 0-D arrays |
| 19 | +#1776 make complex division by zero to yield inf properly |
| 20 | +#1675 add scalar support for the format() function |
| 21 | +#1905 explicitly check for NaNs in allclose() |
| 22 | +#1952 allow floating ddof in std() and var() |
| 23 | +#1948 fix regression for indexing chararrays with empty list |
| 24 | +#2017 fix type hashing |
| 25 | +#2046 deleting array attributes causes segfault |
| 26 | +#2033 a**2.0 has incorrect type |
| 27 | +#2045 make attribute/iterator_element deletions not segfault |
| 28 | +#2021 fix segfault in searchsorted() |
| 29 | +#2073 fix float16 __array_interface__ bug |
| 30 | + |
| 31 | + |
| 32 | +``numpy.lib`` issues fixed |
| 33 | +-------------------------- |
| 34 | + |
| 35 | +#2048 break reference cycle in NpzFile |
| 36 | +#1573 savetxt() now handles complex arrays |
| 37 | +#1387 allow bincount() to accept empty arrays |
| 38 | +#1899 fixed histogramdd() bug with empty inputs |
| 39 | +#1793 fix failing npyio test under py3k |
| 40 | +#1936 fix extra nesting for subarray dtypes |
| 41 | +#1848 make tril/triu return the same dtype as the original array |
| 42 | +#1918 use Py_TYPE to access ob_type, so it works also on Py3 |
| 43 | + |
| 44 | + |
| 45 | +``numpy.f2py`` changes |
| 46 | +---------------------- |
| 47 | + |
| 48 | +ENH: Introduce new options extra_f77_compiler_args and extra_f90_compiler_args |
| 49 | +BLD: Improve reporting of fcompiler value |
| 50 | +BUG: Fix f2py test_kind.py test |
| 51 | + |
| 52 | + |
| 53 | +``numpy.poly`` changes |
| 54 | +---------------------- |
| 55 | + |
| 56 | +ENH: Add some tests for polynomial printing |
| 57 | +ENH: Add companion matrix functions |
| 58 | +DOC: Rearrange the polynomial documents |
| 59 | +BUG: Fix up links to classes |
| 60 | +DOC: Add version added to some of the polynomial package modules |
| 61 | +DOC: Document xxxfit functions in the polynomial package modules |
| 62 | +BUG: The polynomial convenience classes let different types interact |
| 63 | +DOC: Document the use of the polynomial convenience classes |
| 64 | +DOC: Improve numpy reference documentation of polynomial classes |
| 65 | +ENH: Improve the computation of polynomials from roots |
| 66 | +STY: Code cleanup in polynomial [*]fromroots functions |
| 67 | +DOC: Remove references to cast and NA, which were added in 1.7 |
| 68 | + |
| 69 | + |
| 70 | +``numpy.distutils`` issues fixed |
| 71 | +------------------------------- |
| 72 | + |
| 73 | +#1261 change compile flag on AIX from -O5 to -O3 |
| 74 | +#1377 update HP compiler flags |
| 75 | +#1383 provide better support for C++ code on HPUX |
| 76 | +#1857 fix build for py3k + pip |
| 77 | +BLD: raise a clearer warning in case of building without cleaning up first |
| 78 | +BLD: follow build_ext coding convention in build_clib |
| 79 | +BLD: fix up detection of Intel CPU on OS X in system_info.py |
| 80 | +BLD: add support for the new X11 directory structure on Ubuntu & co. |
| 81 | +BLD: add ufsparse to the libraries search path. |
| 82 | +BLD: add 'pgfortran' as a valid compiler in the Portland Group |
| 83 | +BLD: update version match regexp for IBM AIX Fortran compilers. |
| 84 | + |
| 85 | + |
| 86 | +``numpy.random`` issues fixed |
| 87 | +----------------------------- |
| 88 | + |
| 89 | +BUG: Use npy_intp instead of long in mtrand |
| 90 | + |
0 commit comments