Skip to content


Subversion checkout URL

You can clone with
Download ZIP
Commits on May 19, 2012
  1. @rgommers

    REL: set version to 1.6.2.

    rgommers authored
  2. @rgommers

    TST: for numpy.random test, back off test precision a little.

    rgommers authored
    This test was reported to be failing once for 1.6.2rc1, test precision may be
    set slightly too agressive.
  3. @rgommers

    BUG: fix running tests with coverage=True.

    rgommers authored
    The --cover-inclusive argument means that tries to include every
    single .py file in the source tree in the coverage report. This leads to test
    errors, because it tries to import files like (which will of
    course directly fail for anyone not having numscons installed).
    (backport of bfaaefe)
  4. @rgommers
Commits on May 12, 2012
  1. @rgommers

    REL: change release script to generate tarballs before building docs.

    rgommers authored
    This avoids shipping generated .rst files and some .pyc files in the doc/ dir.
  2. @rgommers
Commits on May 5, 2012
  1. @rgommers
  2. @rgommers

    REL: increase allowed size of DMG installer by 1 Mb.

    rgommers authored
    Caused by the docs becoming much larger (~100 pages) due to numpy.poly
Commits on May 4, 2012
  1. @rgommers

    REL: set version to 1.6.2rc1

    rgommers authored
  2. @rgommers
  3. @rgommers

    REL: change Sphinx version in release script to 1.1.3.

    rgommers authored
    Version 1.0.4 was used till now, but has become too old.  New version gives far
    fewer warnings and thereby increases the reference guide with O(100) pages that
    previously got lost.
    Also update current Cython version number.
    (backport of d609dff)
  4. @rgommers

    Merge pull request #265 from charris/backport-random

    rgommers authored
    Backport numpy.random fix for MSVC.
  5. @rgommers

    Merge pull request #263 from charris/backport-poly

    rgommers authored
    Backport numpy.poly bug fixes and documentation improvements.
  6. @rgommers

    Merge pull request #262 from charris/backport-f2py

    rgommers authored
    Backport f2py fixes and improvements.
  7. @rgommers

    Merge pull request #261 from charris/backport-lib

    rgommers authored
    Backport numpy.lib bug fixes.
  8. @rgommers

    Merge pull request #260 from charris/backport-core

    rgommers authored
    Backport numpy.core bug fixes.
    All tested on MSVC, OS X and with a full release build on Windows.
Commits on May 3, 2012
  1. @rgommers @charris

    TST: fix string comparison test failures on Windows for Python 2.5.

    rgommers authored charris committed
    This is caused by the inconsistent floating point handling of Python itself.
    On Windows with 2.5:
    >>> "%s" % 1e-6
    With 2.6:
    >>> "%s" % 1e-6
    Reviewed as PR-225.
  2. @charris
Commits on May 2, 2012
  1. @thouis @charris
Commits on May 1, 2012
  1. @dwf @charris

    BUG: PR #188, PyArray_{Get,Set}Map using low-precision counter

    dwf authored charris committed
    Should be using npy_intp rather than int.
  2. @cgohlke @charris

    BUG: PR #227, Use npy_intp instead of long in mtrand.

    cgohlke authored charris committed
    This fixes a problem on 64 bit Windows where a 64 bit index is
    needed but long is only 32 bits.
    Regenerate mtrand.c
Commits on Apr 30, 2012
  1. @charris
  2. @charris

    STY: Code cleanup in polynomial [*]fromroots functions.

    charris authored
    Use divmod instead of // and % separately.
  3. @charris

    ENH: Improve the computation of polynomials from roots.

    charris authored
    The original method was overly sensitive to roundoff. Of the two
    approaches considered, gauss integration or binary subdivision of
    the roots, the latter is more compatible with using other number
    representations such as mpmath. No method is going to be suitable
    for large numbers of arbitrary zeros but the current method is a
    significant improvement.
  4. @charris

    DOC: Improve numpy reference documentation of polynomial classes.

    charris authored
    Add more explanation of what happens in polynomial substitution and also of
    its use in casting between types. Clean up some rst glitches.
  5. @charris
  6. @charris

    BUG: The polynomial convenience classes let different types interact.

    charris authored
    In particular for arithmetic where one could end up with a Polynomial
    type with Chebyshev coefficients after an addition. It is unlikely that
    that would be done on purpose.
    The PolyDomain error message was also replaced by a TypeError with
    an appropriate message. That seems like a better choice.
  7. @charris
  8. @charris
  9. @charris

    BUG: Fix up links to classes.

    charris authored
    The Sphinx markup was incorrect the documentation to the polynomial
    convenience classes wasn't being generated.
  10. @charris

    DOC: Rearrange the polynomial documents.

    charris authored
    This is the first step in cleaning up the polynomial documentation
    and writing an instructional section on the convenience classes.
  11. @charris

    ENH: Add companion matrix functions.

    charris authored
    The new companion matrices are related to the old by a
    similarity transformation that makes them better conditioned
    for root finding. In particular, the companion matrices for
    the orthogonal polynomials are symmetric when the zeros of a
    single polynomial term is wanted. This produces better zeros
    for use in Gauss quadrature.
  12. @charris
  13. @charris

    BUG: Fix f2py test.

    charris authored
    Newer Fortran compilers for Intel may support quad precision, so
    _selected_real_kind_func needs to report that for precisions >= 19.
  14. @pearu @charris

    ENH: Add new options extra_f77_compile_args and extra_f90_compile_args.

    pearu authored charris committed
    Add new options extra_f77_compile_args and extra_f90_compile_args to
    Configuration.add_extension. Configuration.add_library, and Extension.
    These options allow specifying extra compile options for compiling Fortran
    sources within a file.
Something went wrong with that request. Please try again.