delta/python-packages/numpy.git/numpy/testing, branch v1.8.0b1

REL: Set raise_warnings to "release" in NoseTester constructor.

2013-08-31T20:10:42+00:00

STY: Giant comma spacing fixup.

2013-08-18T17:51:25+00:00

Run the 2to3 ws_comma fixer on *.py files. Some lines are now too long
and will need to be broken at some point. OTOH, some lines were already
too long and need to be broken at some point. Now seems as good a time
as any to do this with open PRs at a minimum.

STY: Giant whitespace cleanup.

2013-08-18T17:20:45+00:00

Now is as good a time as any with open PR's at a low.

MAINT: Refactor nanfunctions.

2013-08-13T04:33:56+00:00

nanmax, nanmin
--------------
Add out and keepdims keywords.

nanargmin, nanargmax
--------------------
A NanWarning is raised if an all NaN slice detected. For all such
slices np.iingo(np.intp).min is returned as the index value.

nansum
------
The keywords dtype, out, and keepdims are added.

A FutureWarning is raised, as in the future the mean of an empty
slice after NaN replacement will be 0 instead of the current NaN.

nanmean, nanvar, nanstd
-----------------------
For all, if the input array is of inexact type then the dtype and out
parameters must be of inexact type if specified.  That insures that NaNs
can be returned when appropriate.

The nanmean function detects empty slices after NaN replacement and
raises a NanWarning. NaN is returned as the value for all such slices.

The nanmean and nanstd functions detect degrees of freedom <= 0 after
NaN replacement and raise a NanWarning. NaN is returned as the value for
all such slices.

MAINT: remove unused and broken parts of numpy.testing

2013-08-10T08:50:36+00:00

Deprecate np.testing.importall - it's pointless and partially broken.

MAINT: Remove uses of the WarningManager class.

2013-07-12T13:27:07+00:00

WarningManager was a workaround for the lack of the with statement
in Python versions < 2.6. As those versions are no longer supported
it can be removed.

Deprecation notes are added to WarningManager and WarningMessage, but
to avoid a cascade of messages in third party apps, no warnings are
raised at this time, that can be done later.

Closes #3519.

MAINT: Use np.errstate context manager.

2013-07-11T22:49:04+00:00

Now that Python < 2.6 is no longer supported we can use the errstate
context manager in places where constructs like
```
    old = seterr(invalid='ignore')
    try:
        blah
    finally:
       seterr(**old)
```
were used.

DEP: Deprecate the oldnumeric and numarray modules.

2013-06-10T22:30:10+00:00

The numarray and oldnumeric modules are deprecated. This is a bit tricky
as raising a DeprecationWarning on import causes an error when tests are
run. To deal with that, a ModuleDeprecationWarning class is added to
numpy and NoseTester is modified to ignore that warning during testing.

Closes #2905

MAINT: fix shape mismatch issue in alignment data generator

2013-06-08T18:44:05+00:00

ENH: vectorize sqrt ufunc using SSE2

2013-05-25T15:36:00+00:00

specialize the sqrt ufunc for float and double and vectorize it using
SSE2.

improves performance by 4/2 for float/double if one is not memory bound
due to non-cached data.
performance is always better on all tested machines (amd phenom X2,
intel xeon 5xxx/7xxx, core2duo, corei7)

This version will not set errno on invalid input, but numpy only checks
the fpu flags so the behavior is the same.

In principle the compiler could autovectorize it when setting ffast-math
(for no errno) and specializing the loop for the vectorizable strides
and giving it some hints (restrict, __builtin_assume_aligned, etc.),
but its simpler and more reliable to simply vectorize it by hand.