Installation#

The easiest way to install pandas is to install it as part of the Anaconda distribution, a cross platform distribution for data analysis and scientific computing. The Conda package manager is the recommended installation method for most users.

Instructions for installing from source, PyPI, or a development version are also provided.

Python version support#

Officially Python 3.9, 3.10 and 3.11.

Installing pandas#

Installing with Anaconda#

For users that are new to Python, the easiest way to install Python, pandas, and the packages that make up the PyData stack (SciPy, NumPy, Matplotlib, and more) is with Anaconda, a cross-platform (Linux, macOS, Windows) Python distribution for data analytics and scientific computing. Installation instructions for Anaconda can be found here.

Installing with Miniconda#

For users experienced with Python, the recommended way to install pandas with Miniconda. Miniconda allows you to create a minimal, self-contained Python installation compared to Anaconda and use the Conda package manager to install additional packages and create a virtual environment for your installation. Installation instructions for Miniconda can be found here.

The next step is to create a new conda environment. A conda environment is like a virtualenv that allows you to specify a specific version of Python and set of libraries. Run the following commands from a terminal window.

conda create -c conda-forge -n name_of_my_env python pandas

This will create a minimal environment with only Python and pandas installed. To put your self inside this environment run.

source activate name_of_my_env
# On Windows
activate name_of_my_env

Installing from PyPI#

pandas can be installed via pip from PyPI.

pip install pandas

Note

You must have pip>=19.3 to install from PyPI.

Note

It is recommended to install and run pandas from a virtual environment, for example, using the Python standard library’s venv

pandas can also be installed with sets of optional dependencies to enable certain functionality. For example, to install pandas with the optional dependencies to read Excel files.

pip install "pandas[excel]"

The full list of extras that can be installed can be found in the dependency section.

Handling ImportErrors#

If you encounter an ImportError, it usually means that Python couldn’t find pandas in the list of available libraries. Python internally has a list of directories it searches through, to find packages. You can obtain these directories with.

import sys
sys.path

One way you could be encountering this error is if you have multiple Python installations on your system and you don’t have pandas installed in the Python installation you’re currently using. In Linux/Mac you can run which python on your terminal and it will tell you which Python installation you’re using. If it’s something like “/usr/bin/python”, you’re using the Python from the system, which is not recommended.

It is highly recommended to use conda, for quick installation and for package and dependency updates. You can find simple installation instructions for pandas in this document.

Installing from source#

See the contributing guide for complete instructions on building from the git source tree. Further, see creating a development environment if you wish to create a pandas development environment.

Installing the development version of pandas#

Installing the development version is the quickest way to:

  • Try a new feature that will be shipped in the next release (that is, a feature from a pull-request that was recently merged to the main branch).

  • Check whether a bug you encountered has been fixed since the last release.

The development version is usually uploaded daily to the scientific-python-nightly-wheels index from the PyPI registry of anaconda.org. You can install it by running.

pip install --pre --extra-index https://pypi.anaconda.org/scientific-python-nightly-wheels/simple pandas

Note that you might be required to uninstall an existing version of pandas to install the development version.

pip uninstall pandas -y

Running the test suite#

pandas is equipped with an exhaustive set of unit tests. The packages required to run the tests can be installed with pip install "pandas[test]". To run the tests from a Python terminal.

>>> import pandas as pd
>>> pd.test()
running: pytest -m "not slow and not network and not db" /home/user/anaconda3/lib/python3.9/site-packages/pandas

============================= test session starts ==============================
platform linux -- Python 3.9.7, pytest-6.2.5, py-1.11.0, pluggy-1.0.0
rootdir: /home/user
plugins: dash-1.19.0, anyio-3.5.0, hypothesis-6.29.3
collected 154975 items / 4 skipped / 154971 selected
........................................................................ [  0%]
........................................................................ [ 99%]
.......................................                                  [100%]

==================================== ERRORS ====================================

=================================== FAILURES ===================================

=============================== warnings summary ===============================

=========================== short test summary info ============================

= 1 failed, 146194 passed, 7402 skipped, 1367 xfailed, 5 xpassed, 197 warnings, 10 errors in 1090.16s (0:18:10) =

Note

This is just an example of what information is shown. Test failures are not necessarily indicative of a broken pandas installation.

Dependencies#

Required dependencies#

pandas requires the following dependencies.

Package

Minimum supported version

NumPy

1.22.4

python-dateutil

2.8.2

pytz

2020.1

Optional dependencies#

pandas has many optional dependencies that are only used for specific methods. For example, pandas.read_hdf() requires the pytables package, while DataFrame.to_markdown() requires the tabulate package. If the optional dependency is not installed, pandas will raise an ImportError when the method requiring that dependency is called.

If using pip, optional pandas dependencies can be installed or managed in a file (e.g. requirements.txt or pyproject.toml) as optional extras (e.g. pandas[performance, aws]). All optional dependencies can be installed with pandas[all], and specific sets of dependencies are listed in the sections below.

Visualization#

Installable with pip install "pandas[plot, output_formatting]".

Dependency

Minimum Version

pip extra

Notes

matplotlib

3.6.1

plot

Plotting library

Jinja2

3.1.2

output_formatting

Conditional formatting with DataFrame.style

tabulate

0.8.10

output_formatting

Printing in Markdown-friendly format (see tabulate)

Computation#

Installable with pip install "pandas[computation]".

Dependency

Minimum Version

pip extra

Notes

SciPy

1.8.1

computation

Miscellaneous statistical functions

xarray

2022.03.0

computation

pandas-like API for N-dimensional data

Excel files#

Installable with pip install "pandas[excel]".

Dependency

Minimum Version

pip extra

Notes

xlrd

2.0.1

excel

Reading Excel

xlsxwriter

3.0.3

excel

Writing Excel

openpyxl

3.0.10

excel

Reading / writing for xlsx files

pyxlsb

1.0.9

excel

Reading for xlsb files

HTML#

Installable with pip install "pandas[html]".

Dependency

Minimum Version

pip extra

Notes

BeautifulSoup4

4.11.1

html

HTML parser for read_html

html5lib

1.1

html

HTML parser for read_html

lxml

4.8.0

html

HTML parser for read_html

One of the following combinations of libraries is needed to use the top-level read_html() function:

Warning

XML#

Installable with pip install "pandas[xml]".

Dependency

Minimum Version

pip extra

Notes

lxml

4.8.0

xml

XML parser for read_xml and tree builder for to_xml

SQL databases#

Installable with pip install "pandas[postgresql, mysql, sql-other]".

Dependency

Minimum Version

pip extra

Notes

SQLAlchemy

1.4.36

postgresql, mysql, sql-other

SQL support for databases other than sqlite

psycopg2

2.9.3

postgresql

PostgreSQL engine for sqlalchemy

pymysql

1.0.2

mysql

MySQL engine for sqlalchemy

Other data sources#

Installable with pip install "pandas[hdf5, parquet, feather, spss, excel]"

Dependency

Minimum Version

pip extra

Notes

PyTables

3.7.0

hdf5

HDF5-based reading / writing

blosc

1.21.0

hdf5

Compression for HDF5; only available on conda

zlib

hdf5

Compression for HDF5

fastparquet

0.8.1

Parquet reading / writing (pyarrow is default)

pyarrow

7.0.0

parquet, feather

Parquet, ORC, and feather reading / writing

pyreadstat

1.1.5

spss

SPSS files (.sav) reading

odfpy

1.4.1

excel

Open document format (.odf, .ods, .odt) reading / writing

Warning

  • If you want to use read_orc(), it is highly recommended to install pyarrow using conda. read_orc() may fail if pyarrow was installed from pypi, and read_orc() is not compatible with Windows OS.

Access data in the cloud#

Installable with pip install "pandas[fss, aws, gcp]"

Dependency

Minimum Version

pip extra

Notes

fsspec

2022.05.0

fss, gcp, aws

Handling files aside from simple local and HTTP (required dependency of s3fs, gcsfs).

gcsfs

2022.05.0

gcp

Google Cloud Storage access

pandas-gbq

0.17.5

gcp

Google Big Query access

s3fs

2022.05.0

aws

Amazon S3 access

Clipboard#

Installable with pip install "pandas[clipboard]".

Dependency

Minimum Version

pip extra

Notes

PyQt4/PyQt5

5.15.6

clipboard

Clipboard I/O

qtpy

2.2.0

clipboard

Clipboard I/O

Note

Depending on operating system, system-level packages may need to installed. For clipboard to operate on Linux one of the CLI tools xclip or xsel must be installed on your system.

Compression#

Installable with pip install "pandas[compression]"

Dependency

Minimum Version

pip extra

Notes

brotli

0.7.0

compression

Brotli compression

python-snappy

0.6.1

compression

Snappy compression

Zstandard

0.17.0

compression

Zstandard compression

Consortium Standard#

Installable with pip install "pandas[consortium-standard]"

Dependency

Minimum Version

pip extra

Notes

dataframe-api-compat

0.1.7

consortium-standard

Consortium Standard-compatible implementation based on pandas