Package overview

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 thefundamental high-level building block for doing practical, real world dataanalysis in Python. Additionally, it has the broader goal of becoming themost powerful and flexible open source data analysis / manipulation toolavailable in any language. It is already well on its way toward this goal.

pandas is well suited for many different kinds of data:

  • Tabular data with heterogeneously-typed columns, as in an SQL table orExcel spreadsheet
  • Ordered and unordered (not necessarily fixed-frequency) time series data.
  • Arbitrary matrix data (homogeneously typed or heterogeneous) with row andcolumn labels
  • Any other form of observational / statistical data sets. The data actuallyneed not be labeled at all to be placed into a pandas data structure

The two primary data structures of pandas, Series (1-dimensional)and DataFrame (2-dimensional), handle the vast majority of typical usecases in finance, statistics, social science, and many areas ofengineering. For R users, DataFrame provides everything that R’sdata.frame provides and much more. pandas is built on top of NumPy and is intended to integrate well within a scientificcomputing environment with many other 3rd party libraries.

Here are just a few of the things that pandas does well:

  • Easy handling of missing data (represented as NaN) in floating point aswell as non-floating point data
  • Size mutability: columns can be inserted and deleted from DataFrame andhigher dimensional objects
  • Automatic and explicit data alignment: objects can be explicitlyaligned to a set of labels, or the user can simply ignore the labels andlet Series, DataFrame, etc. automatically align the data for you incomputations
  • Powerful, flexible group by functionality to performsplit-apply-combine operations on data sets, for both aggregating andtransforming data
  • Make it easy to convert ragged, differently-indexed data in otherPython and NumPy data structures into DataFrame objects
  • Intelligent label-based slicing, fancy indexing, and subsettingof large data sets
  • Intuitive merging and joining data sets
  • Flexible reshaping and pivoting of data sets
  • Hierarchical labeling of axes (possible to have multiple labels pertick)
  • Robust IO tools for loading data from flat files (CSV and delimited),Excel files, databases, and saving / loading data from the ultrafast HDF5format
  • Time series-specific functionality: date range generation and frequencyconversion, moving window statistics, moving window linear regressions,date shifting and lagging, etc.

Many of these principles are here to address the shortcomings frequentlyexperienced using other languages / scientific research environments. For datascientists, working with data is typically divided into multiple stages:munging and cleaning data, analyzing / modeling it, then organizing the resultsof the analysis into a form suitable for plotting or tabular display. pandasis the ideal tool for all of these tasks.

Some other notes

  • pandas is fast. Many of the low-level algorithmic bits have beenextensively tweaked in Cython code. However, as withanything else generalization usually sacrifices performance. So if you focuson one feature for your application you may be able to create a fasterspecialized tool.
  • pandas is a dependency of statsmodels, making it an important part of thestatistical computing ecosystem in Python.
  • pandas has been used extensively in production in financial applications.

Data structures

DimensionsNameDescription
1Series1D labeled homogeneously-typed array
2DataFrameGeneral 2D labeled, size-mutable tabular structure with potentially heterogeneously-typed column

Why more than one data structure?

The best way to think about the pandas data structures is as flexiblecontainers for lower dimensional data. For example, DataFrame is a containerfor Series, and Series is a container for scalars. We would like to beable to insert and remove objects from these containers in a dictionary-likefashion.

Also, we would like sensible default behaviors for the common API functionswhich take into account the typical orientation of time series andcross-sectional data sets. When using ndarrays to store 2- and 3-dimensionaldata, a burden is placed on the user to consider the orientation of the dataset when writing functions; axes are considered more or less equivalent (exceptwhen C- or Fortran-contiguousness matters for performance). In pandas, the axesare intended to lend more semantic meaning to the data; i.e., for a particulardata 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 datatransformations in downstream functions.

For example, with tabular data (DataFrame) it is more semantically helpful tothink of the index (the rows) and the columns rather than axis 0 andaxis 1. Iterating through the columns of the DataFrame thus results in morereadable code:

  1. for col in df.columns:
  2.     series = df[col]
  3.     # do something with series

Mutability and copying of data

All pandas data structures are value-mutable (the values they contain can bealtered) but not always size-mutable. The length of a Series cannot bechanged, but, for example, columns can be inserted into a DataFrame. However,the vast majority of methods produce new objects and leave the input datauntouched. In general we like to favor immutability where sensible.

Getting support

The first stop for pandas issues and ideas is the Github Issue Tracker. If you have a general question,pandas community experts can answer through Stack Overflow.

Community

pandas is actively supported today by a community of like-minded individuals aroundthe world who contribute their valuable time and energy to help make open sourcepandas possible. Thanks to all of our contributors.

If you’re interested in contributing, please visit the contributing guide.

pandas is a NumFOCUS sponsored project.This will help ensure the success of development of pandas as a world-class open-sourceproject, and makes it possible to donate to the project.

Project governance

The governance process that pandas project has used informally since its inception in 2008 is formalized in Project Governance documents.The documents clarify how decisions are made and how the various elements of our community interact, including the relationship between open source collaborative development and work that may be funded by for-profit or non-profit entities.

Wes McKinney is the Benevolent Dictator for Life (BDFL).

Development team

The list of the Core Team members and more detailed information can be found on the people’s page of the governance repo.

Institutional partners

The information about current institutional partners can be found on pandas website page.

License

  1. BSD 3-Clause License
  2.  
  3. Copyright (c) 2008-2012, AQR Capital Management, LLC, Lambda Foundry, Inc. and PyData Development Team
  4. All rights reserved.
  5.  
  6. Redistribution and use in source and binary forms, with or without
  7. modification, are permitted provided that the following conditions are met:
  8.  
  9. * Redistributions of source code must retain the above copyright notice, this
  10.   list of conditions and the following disclaimer.
  11.  
  12. * Redistributions in binary form must reproduce the above copyright notice,
  13.   this list of conditions and the following disclaimer in the documentation
  14.   and/or other materials provided with the distribution.
  15.  
  16. * Neither the name of the copyright holder nor the names of its
  17.   contributors may be used to endorse or promote products derived from
  18.   this software without specific prior written permission.
  19.  
  20. THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
  21. AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  22. IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
  23. DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
  24. FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  25. DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
  26. SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
  27. CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
  28. OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
  29. OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.