pandas consists of the following things
- A set of labeled array data structures, the primary of which are Series/TimeSeries and DataFrame
- Index objects enabling both simple axis indexing and multi-level / hierarchical axis indexing
- An integrated group by engine for aggregating and transforming data sets
- Date range generation (date_range) and custom date offsets enabling the implementation of customized frequencies
- Input/Output tools: loading tabular data from flat files (CSV, delimited, Excel 2003), and saving and loading pandas objects from the fast and efficient PyTables/HDF5 format.
- Memory-efficient “sparse” versions of the standard data structures for storing data that is mostly missing or mostly constant (some fixed value)
- Moving window statistics (rolling mean, rolling standard deviation, etc.)
- Static and moving window linear and panel regression
Data structures at a glance¶
|1||Series||1D labeled homogeneously-typed array|
|1||TimeSeries||Series with index containing datetimes|
|2||DataFrame||General 2D labeled, size-mutable tabular structure with potentially heterogeneously-typed columns|
|3||Panel||General 3D labeled, also size-mutable array|
Why more than 1 data structure?¶
The best way to think about the pandas data structures is as flexible containers for lower dimensional data. For example, DataFrame is a container for Series, and Panel is a container for DataFrame objects. We would like to be able to insert and remove objects from these containers in a dictionary-like fashion.
Also, we would like sensible default behaviors for the common API functions which take into account the typical orientation of time series and cross-sectional data sets. When using ndarrays to store 2- and 3-dimensional data, a burden is placed on the user to consider the orientation of the data set when writing functions; axes are considered more or less equivalent (except when C- or Fortran-contiguousness matters for performance). In pandas, the axes are intended to lend more semantic meaning to the data; i.e., for a particular data set there is likely to be a “right” way to orient the data. The goal, then, is to reduce the amount of mental effort required to code up data transformations in downstream functions.
For example, with tabular data (DataFrame) it is more semantically helpful to think of the index (the rows) and the columns rather than axis 0 and axis 1. And iterating through the columns of the DataFrame thus results in more readable code:
for col in df.columns: series = df[col] # do something with series
Mutability and copying of data¶
All pandas data structures are value-mutable (the values they contain can be altered) but not always size-mutable. The length of a Series cannot be changed, but, for example, columns can be inserted into a DataFrame. However, the vast majority of methods produce new objects and leave the input data untouched. In general, though, we like to favor immutability where sensible.
pandas development began at AQR Capital Management in April 2008. It was open-sourced at the end of 2009. AQR continued to provide resources for development through the end of 2011, and continues to contribute bug reports today.
Since January 2012, Lambda Foundry, has been providing development resources, as well as commercial support, training, and consulting for pandas.
pandas is only made possible by a group of people around the world like you who have contributed new code, bug reports, fixes, comments and ideas. A complete list can be found on Github.
pandas is a part of the PyData project. The PyData Development Team is a collection of developers focused on the improvement of Python’s data libraries. The core team that coordinates development can be found on Github. If you’re interested in contributing, please visit the project website.
======= License ======= pandas is distributed under a 3-clause ("Simplified" or "New") BSD license. Parts of NumPy, SciPy, numpydoc, bottleneck, which all have BSD-compatible licenses, are included. Their licenses follow the pandas license. pandas license ============== Copyright (c) 2011-2012, Lambda Foundry, Inc. and PyData Development Team All rights reserved. Copyright (c) 2008-2011 AQR Capital Management, LLC All rights reserved. Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. * Neither the name of the copyright holder nor the names of any contributors may be used to endorse or promote products derived from this software without specific prior written permission. THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. About the Copyright Holders =========================== AQR Capital Management began pandas development in 2008. Development was led by Wes McKinney. AQR released the source under this license in 2009. Wes is now an employee of Lambda Foundry, and remains the pandas project lead. The PyData Development Team is the collection of developers of the PyData project. This includes all of the PyData sub-projects, including pandas. The core team that coordinates development on GitHub can be found here: http://github.com/pydata. Full credits for pandas contributors can be found in the documentation. Our Copyright Policy ==================== PyData uses a shared copyright model. Each contributor maintains copyright over their contributions to PyData. However, it is important to note that these contributions are typically only changes to the repositories. Thus, the PyData source code, in its entirety, is not the copyright of any single person or institution. Instead, it is the collective copyright of the entire PyData Development Team. If individual contributors want to maintain a record of what changes/contributions they have specific copyright on, they should indicate their copyright in the commit message of the change when they commit the change to one of the PyData repositories. With this in mind, the following banner should be used in any source code file to indicate the copyright and license terms: #----------------------------------------------------------------------------- # Copyright (c) 2012, PyData Development Team # All rights reserved. # # Distributed under the terms of the BSD Simplified License. # # The full license is in the LICENSE file, distributed with this software. #----------------------------------------------------------------------------- Other licenses can be found in the LICENSES directory.