[SciPy-User] ANN: pandas 0.7.0 release
Thu Feb 9 17:13:29 CST 2012
I'm extremely pleased (and relieved!) to announce the release of
pandas 0.7.0! This is the largest single release in the last
year, spanning 210 GitHub issues and pull requests with 563
commits from 17 unique authors. It brings with it a wealth of new
functionality, performance improvements, bug fixes, and a handful
of very minor API changes. I recommend that all users upgrade to
the new release as soon as you can.
Here are a few highlights of the release:
* Completely revamped, high-performance merge/join
infrastructure. Full support for all SQL-style joins. Fastest
open source implementation I am aware of.
* New unified concat function for easily concatenating pandas objects
* Better pivot table and cross-tabulation functionality
* Numerous new Series and DataFrame instance methods
* Substantially improved performance of GroupBy operations
* Excel 2007 read/write support
* Much better unicode handling on both Python 2 and 3
* Improved console DataFrame formatting
* More than 70 bug fixes
* Numerous other performance and infrastructural improvements
This release also coincides with the creation of a new tool,
vbench (https://github.com/wesm/vbench), for systematically
monitoring the performance of Python code (in this case pandas)
over time. There are now 57 vbenchmarks being tracked with more
added all the time (http://pandas.pydata.org/pandas-docs/vbench/).
This will help ensure that pandas remains a high performance library
in addition to being robust and stable for production application
pandas has a new project front page at http://pandas.pydata.org.
The main repository has also been moved to the newly created
PyData organization on GitHub (http://github.com/pydata).
Windows binaries are available on PyPI, and .deb binaries will be
available in Debian sid and NeuroDebian soon thanks to Yaroslav
See the "What's New" page and full release notes for now. Thanks
to everyone who contributed to the release! Tons more planned for
pandas in 2012 on the road toward a 1.0 release. Looking forward
to working with the community to make the library even better!
What is it
pandas is a Python package providing fast, flexible, and expressive
data structures designed to make working with “relational” or
“labeled” data both easy and intuitive. It aims to be the fundamental
high-level building block for doing practical, real world data
analysis in Python.
Release Notes: http://github.com/pydata/pandas/blob/master/RELEASE.rst
Code Repository: http://github.com/pydata/pandas
Mailing List: http://groups.google.com/group/pystatsmodels
More information about the SciPy-User