Merge pull request #7994 from jreback/pytables_api

COMPAT: change pytables to use 3.0.0 API (GH7990)

Author: jreback

ci/requirements-2.7.txt

@@ -5,7 +5,7 @@ numpy==1.8.1
 cython==0.19.1
 bottleneck==0.6.0
 numexpr==2.2.2
-tables==2.3.1
+tables==3.0.0
 matplotlib==1.3.1
 openpyxl==1.6.2
 xlsxwriter==0.4.6

doc/source/install.rst

@@ -26,7 +26,7 @@ Installing pandas
 Trying out pandas, no installation required!
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-The easiest way to start experimenting with pandas doesn't involve installing 
+The easiest way to start experimenting with pandas doesn't involve installing
 pandas at all.
 
 `Wakari <https://wakari.io>`__ is a free service that provides a hosted
@@ -35,10 +35,10 @@ pandas at all.
 Simply create an account, and have access to pandas from within your brower via
 an `IPython Notebook <http://ipython.org/notebook.html>`__ in a few minutes.
 
-Installing pandas with Anaconda 
+Installing pandas with Anaconda
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-Installing pandas and the rest of the `NumPy <http://www.numpy.org/>`__ and 
+Installing pandas and the rest of the `NumPy <http://www.numpy.org/>`__ and
 `SciPy <http://www.scipy.org/>`__ stack can be a little
 difficult for inexperienced users.
 
@@ -57,8 +57,8 @@ anything else, and without needing to wait for any software to be compiled.
 Installation instructions for `Anaconda <http://docs.continuum.io/anaconda/>`__
 `can be found here <http://docs.continuum.io/anaconda/install.html>`__.
 
-A full list of the packages available as part of the 
-`Anaconda <http://docs.continuum.io/anaconda/>`__ distribution 
+A full list of the packages available as part of the
+`Anaconda <http://docs.continuum.io/anaconda/>`__ distribution
 `can be found here <http://docs.continuum.io/anaconda/pkg-docs.html>`__.
 
 An additional advantage of installing with Anaconda is that you don't require
@@ -78,7 +78,7 @@ If you want to have more control on which packages, or have a limited internet
 bandwidth, then installing pandas with
 `Miniconda <http://conda.pydata.org/miniconda.html>`__ may be a better solution.
 
-`Conda <http://conda.pydata.org/docs/>`__ is the package manager that the 
+`Conda <http://conda.pydata.org/docs/>`__ is the package manager that the
 `Anaconda <http://docs.continuum.io/anaconda/>`__ distribution is built upon.
 It is a package manager that is both cross-platform and language agnostic
 (it can play a similar role to a pip and virtualenv combination).
@@ -90,15 +90,15 @@ minimal self contained Python installation, and then use the
 First you will need `Conda <http://conda.pydata.org/docs/>`__ to be installed and
 downloading and running the `Miniconda
 <http://conda.pydata.org/miniconda.html>`__
-will do this for you. The installer 
+will do this for you. The installer
 `can be found here <http://conda.pydata.org/miniconda.html>`__
 
 The next step is to create a new conda environment (these are analogous to a
 virtualenv but they also allow you to specify precisely which Python version
 to install also). Run the following commands from a terminal window::
 
     conda create -n name_of_my_env python
-    
+
 This will create a minimal environment with only Python installed in it.
 To put your self inside this environment run::
 
@@ -108,7 +108,7 @@ On Windows the command is::
 
     activate name_of_my_env
 
-The final step required is to install pandas. This can be done with the 
+The final step required is to install pandas. This can be done with the
 following command::
 
     conda install pandas
@@ -143,7 +143,7 @@ pandas can be installed via pip from
     pip install pandas
 
 This will likely require the installation of a number of dependencies,
-including NumPy, will require a compiler to compile required bits of code, 
+including NumPy, will require a compiler to compile required bits of code,
 and can take a few minutes to complete.
 
 Installing using your Linux distribution's package manager.
@@ -259,6 +259,7 @@ Recommended Dependencies
 
   * `numexpr <http://code.google.com/p/numexpr/>`__: for accelerating certain numerical operations.
     ``numexpr`` uses multiple cores as well as smart chunking and caching to achieve large speedups.
+    If installed, must be Version 2.1 or higher.
 
   * `bottleneck <http://berkeleyanalytics.com/bottleneck>`__: for accelerating certain types of ``nan``
     evaluations. ``bottleneck`` uses specialized cython routines to achieve large speedups.
@@ -277,7 +278,7 @@ Optional Dependencies
   * `Cython <http://www.cython.org>`__: Only necessary to build development
     version. Version 0.17.1 or higher.
   * `SciPy <http://www.scipy.org>`__: miscellaneous statistical functions
-  * `PyTables <http://www.pytables.org>`__: necessary for HDF5-based storage
+  * `PyTables <http://www.pytables.org>`__: necessary for HDF5-based storage. Version 3.0.0 or higher required.
   * `SQLAlchemy <http://www.sqlalchemy.org>`__: for SQL database support. Version 0.8.1 or higher recommended.
   * `matplotlib <http://matplotlib.sourceforge.net/>`__: for plotting
   * `statsmodels <http://statsmodels.sourceforge.net/>`__

doc/source/io.rst

@@ -2199,12 +2199,9 @@ the high performance HDF5 format using the excellent `PyTables
 <http://www.pytables.org/>`__ library. See the :ref:`cookbook <cookbook.hdf>`
 for some advanced strategies
 
-.. note::
+.. warning::
 
-   ``PyTables`` 3.0.0 was recently released to enable support for Python 3.
-   pandas should be fully compatible (and previously written stores should be
-   backwards compatible) with all ``PyTables`` >= 2.3. For ``python >= 3.2``,
-   ``pandas >= 0.12.0`` is required for compatibility.
+   As of version 0.15.0, pandas requires ``PyTables`` >= 3.0.0. Stores written with prior versions of pandas / ``PyTables`` >= 2.3 are fully compatible (this was the previous minimum ``PyTables`` required version).
 
 .. ipython:: python
    :suppress:

doc/source/v0.15.0.txt

@@ -11,7 +11,8 @@ users upgrade to this version.
 
   - The ``Categorical`` type was integrated as a first-class pandas type, see :ref:`here <whatsnew_0150.cat>`
   - Internal refactoring of the ``Index`` class to no longer sub-class ``ndarray``, see :ref:`Internal Refactoring <whatsnew_0150.refactoring>`
-  - New datetimelike properties accessor ``.dt`` for Series, see :ref:`Dateimelike Properties <whatsnew_0150.dt>`
+  - New datetimelike properties accessor ``.dt`` for Series, see :ref:`Datetimelike Properties <whatsnew_0150.dt>`
+  - dropping support for ``PyTables`` less than version 3.0.0, and ``numexpr`` less than version 2.1 (:issue:`7990`)
 
 - :ref:`Other Enhancements <whatsnew_0150.enhancements>`
 

pandas/computation/eval.py

@@ -42,9 +42,9 @@ def _check_engine(engine):
                               "if 'numexpr' is not installed")
         else:
             ne_version = numexpr.__version__
-            if ne_version < LooseVersion('2.0'):
+            if ne_version < LooseVersion('2.1'):
                 raise ImportError("'numexpr' version is %s, "
-                                  "must be >= 2.0" % ne_version)
+                                  "must be >= 2.1" % ne_version)
 
 
 def _check_parser(parser):

pandas/computation/expressions.py

@@ -13,7 +13,12 @@
 
 try:
     import numexpr as ne
-    _NUMEXPR_INSTALLED = ne.__version__ >= LooseVersion('2.0')
+    ver = ne.__version__
+    _NUMEXPR_INSTALLED = ver >= LooseVersion('2.1')
+    if not _NUMEXPR_INSTALLED:
+        warnings.warn("The installed version of numexpr {ver} is not supported "
+                      "in pandas and will be not be used".format(ver=ver), UserWarning)
+
 except ImportError:  # pragma: no cover
     _NUMEXPR_INSTALLED = False
 

pandas/computation/tests/test_eval.py

@@ -81,7 +81,6 @@ def _is_py3_complex_incompat(result, expected):
 
 _good_arith_ops = com.difference(_arith_ops_syms, _special_case_arith_ops_syms)
 
-
 class TestEvalNumexprPandas(tm.TestCase):
 
     @classmethod
@@ -1515,9 +1514,9 @@ def testit():
         except ImportError:
             raise nose.SkipTest("no numexpr")
         else:
-            if ne.__version__ < LooseVersion('2.0'):
+            if ne.__version__ < LooseVersion('2.1'):
                 with tm.assertRaisesRegexp(ImportError, "'numexpr' version is "
-                                           ".+, must be >= 2.0"):
+                                           ".+, must be >= 2.1"):
                     testit()
             else:
                 testit()