1 .. highlightlang:: none 2 3 .. _installing-index: 4 5 ************************* 6 Installing Python Modules 7 ************************* 8 9 :Email: distutils-sig (a] python.org 10 11 As a popular open source development project, Python has an active 12 supporting community of contributors and users that also make their software 13 available for other Python developers to use under open source license terms. 14 15 This allows Python users to share and collaborate effectively, benefiting 16 from the solutions others have already created to common (and sometimes 17 even rare!) problems, as well as potentially contributing their own 18 solutions to the common pool. 19 20 This guide covers the installation part of the process. For a guide to 21 creating and sharing your own Python projects, refer to the 22 :ref:`distribution guide <distributing-index>`. 23 24 .. note:: 25 26 For corporate and other institutional users, be aware that many 27 organisations have their own policies around using and contributing to 28 open source software. Please take such policies into account when making 29 use of the distribution and installation tools provided with Python. 30 31 32 Key terms 33 ========= 34 35 * ``pip`` is the preferred installer program. Starting with Python 3.4, it 36 is included by default with the Python binary installers. 37 * A *virtual environment* is a semi-isolated Python environment that allows 38 packages to be installed for use by a particular application, rather than 39 being installed system wide. 40 * ``venv`` is the standard tool for creating virtual environments, and has 41 been part of Python since Python 3.3. Starting with Python 3.4, it 42 defaults to installing ``pip`` into all created virtual environments. 43 * ``virtualenv`` is a third party alternative (and predecessor) to 44 ``venv``. It allows virtual environments to be used on versions of 45 Python prior to 3.4, which either don't provide ``venv`` at all, or 46 aren't able to automatically install ``pip`` into created environments. 47 * The `Python Packaging Index <https://pypi.org>`__ is a public 48 repository of open source licensed packages made available for use by 49 other Python users. 50 * the `Python Packaging Authority 51 <https://www.pypa.io/>`__ is the group of 52 developers and documentation authors responsible for the maintenance and 53 evolution of the standard packaging tools and the associated metadata and 54 file format standards. They maintain a variety of tools, documentation, 55 and issue trackers on both `GitHub <https://github.com/pypa>`__ and 56 `BitBucket <https://bitbucket.org/pypa/>`__. 57 * ``distutils`` is the original build and distribution system first added to 58 the Python standard library in 1998. While direct use of ``distutils`` is 59 being phased out, it still laid the foundation for the current packaging 60 and distribution infrastructure, and it not only remains part of the 61 standard library, but its name lives on in other ways (such as the name 62 of the mailing list used to coordinate Python packaging standards 63 development). 64 65 .. deprecated:: 3.6 66 ``pyvenv`` was the recommended tool for creating virtual environments for 67 Python 3.3 and 3.4, and is `deprecated in Python 3.6 68 <https://docs.python.org/dev/whatsnew/3.6.html#deprecated-features>`_. 69 70 .. versionchanged:: 3.5 71 The use of ``venv`` is now recommended for creating virtual environments. 72 73 .. seealso:: 74 75 `Python Packaging User Guide: Creating and using virtual environments 76 <https://packaging.python.org/installing/#creating-virtual-environments>`__ 77 78 79 Basic usage 80 =========== 81 82 The standard packaging tools are all designed to be used from the command 83 line. 84 85 The following command will install the latest version of a module and its 86 dependencies from the Python Packaging Index:: 87 88 python -m pip install SomePackage 89 90 .. note:: 91 92 For POSIX users (including Mac OS X and Linux users), the examples in 93 this guide assume the use of a :term:`virtual environment`. 94 95 For Windows users, the examples in this guide assume that the option to 96 adjust the system PATH environment variable was selected when installing 97 Python. 98 99 It's also possible to specify an exact or minimum version directly on the 100 command line. When using comparator operators such as ``>``, ``<`` or some other 101 special character which get interpreted by shell, the package name and the 102 version should be enclosed within double quotes:: 103 104 python -m pip install SomePackage==1.0.4 # specific version 105 python -m pip install "SomePackage>=1.0.4" # minimum version 106 107 Normally, if a suitable module is already installed, attempting to install 108 it again will have no effect. Upgrading existing modules must be requested 109 explicitly:: 110 111 python -m pip install --upgrade SomePackage 112 113 More information and resources regarding ``pip`` and its capabilities can be 114 found in the `Python Packaging User Guide <https://packaging.python.org>`__. 115 116 Creation of virtual environments is done through the :mod:`venv` module. 117 Installing packages into an active virtual environment uses the commands shown 118 above. 119 120 .. seealso:: 121 122 `Python Packaging User Guide: Installing Python Distribution Packages 123 <https://packaging.python.org/installing/>`__ 124 125 126 How do I ...? 127 ============= 128 129 These are quick answers or links for some common tasks. 130 131 ... install ``pip`` in versions of Python prior to Python 3.4? 132 -------------------------------------------------------------- 133 134 Python only started bundling ``pip`` with Python 3.4. For earlier versions, 135 ``pip`` needs to be "bootstrapped" as described in the Python Packaging 136 User Guide. 137 138 .. seealso:: 139 140 `Python Packaging User Guide: Requirements for Installing Packages 141 <https://packaging.python.org/installing/#requirements-for-installing-packages>`__ 142 143 144 .. installing-per-user-installation: 145 146 ... install packages just for the current user? 147 ----------------------------------------------- 148 149 Passing the ``--user`` option to ``python -m pip install`` will install a 150 package just for the current user, rather than for all users of the system. 151 152 153 ... install scientific Python packages? 154 --------------------------------------- 155 156 A number of scientific Python packages have complex binary dependencies, and 157 aren't currently easy to install using ``pip`` directly. At this point in 158 time, it will often be easier for users to install these packages by 159 `other means <https://packaging.python.org/science/>`__ 160 rather than attempting to install them with ``pip``. 161 162 .. seealso:: 163 164 `Python Packaging User Guide: Installing Scientific Packages 165 <https://packaging.python.org/science/>`__ 166 167 168 ... work with multiple versions of Python installed in parallel? 169 ---------------------------------------------------------------- 170 171 On Linux, Mac OS X, and other POSIX systems, use the versioned Python commands 172 in combination with the ``-m`` switch to run the appropriate copy of 173 ``pip``:: 174 175 python2 -m pip install SomePackage # default Python 2 176 python2.7 -m pip install SomePackage # specifically Python 2.7 177 python3 -m pip install SomePackage # default Python 3 178 python3.4 -m pip install SomePackage # specifically Python 3.4 179 180 Appropriately versioned ``pip`` commands may also be available. 181 182 On Windows, use the ``py`` Python launcher in combination with the ``-m`` 183 switch:: 184 185 py -2 -m pip install SomePackage # default Python 2 186 py -2.7 -m pip install SomePackage # specifically Python 2.7 187 py -3 -m pip install SomePackage # default Python 3 188 py -3.4 -m pip install SomePackage # specifically Python 3.4 189 190 .. other questions: 191 192 Once the Development & Deployment part of PPUG is fleshed out, some of 193 those sections should be linked from new questions here (most notably, 194 we should have a question about avoiding depending on PyPI that links to 195 https://packaging.python.org/en/latest/mirrors/) 196 197 198 Common installation issues 199 ========================== 200 201 Installing into the system Python on Linux 202 ------------------------------------------ 203 204 On Linux systems, a Python installation will typically be included as part 205 of the distribution. Installing into this Python installation requires 206 root access to the system, and may interfere with the operation of the 207 system package manager and other components of the system if a component 208 is unexpectedly upgraded using ``pip``. 209 210 On such systems, it is often better to use a virtual environment or a 211 per-user installation when installing packages with ``pip``. 212 213 214 Pip not installed 215 ----------------- 216 217 It is possible that ``pip`` does not get installed by default. One potential fix is:: 218 219 python -m ensurepip --default-pip 220 221 There are also additional resources for `installing pip. 222 <https://packaging.python.org/tutorials/installing-packages/#install-pip-setuptools-and-wheel>`__ 223 224 225 Installing binary extensions 226 ---------------------------- 227 228 Python has typically relied heavily on source based distribution, with end 229 users being expected to compile extension modules from source as part of 230 the installation process. 231 232 With the introduction of support for the binary ``wheel`` format, and the 233 ability to publish wheels for at least Windows and Mac OS X through the 234 Python Packaging Index, this problem is expected to diminish over time, 235 as users are more regularly able to install pre-built extensions rather 236 than needing to build them themselves. 237 238 Some of the solutions for installing `scientific software 239 <https://packaging.python.org/science/>`__ 240 that are not yet available as pre-built ``wheel`` files may also help with 241 obtaining other binary extensions without needing to build them locally. 242 243 .. seealso:: 244 245 `Python Packaging User Guide: Binary Extensions 246 <https://packaging.python.org/extensions/>`__ 247
