pip install -r requirements.txt fails with the exception below OSError: [Errno 13] Permission denied: '/usr/local/lib/.... What’s wrong and how do I fix this? (I am trying to setup Django)

Installing collected packages: amqp, anyjson, arrow, beautifulsoup4, billiard, boto, braintree, celery, cffi, cryptography, Django, django-bower, django-braces, django-celery, django-crispy-forms, django-debug-toolbar, django-disqus, django-embed-video, django-filter, django-merchant, django-pagination, django-payments, django-storages, django-vote, django-wysiwyg-redactor, easy-thumbnails, enum34, gnureadline, idna, ipaddress, ipython, kombu, mock, names, ndg-httpsclient, Pillow, pyasn1, pycparser, pycrypto, PyJWT, pyOpenSSL, python-dateutil, pytz, requests, six, sqlparse, stripe, suds-jurko
Cleaning up...
Exception:
Traceback (most recent call last):
  File "/usr/lib/python2.7/dist-packages/pip/basecommand.py", line 122, in main
    status = self.run(options, args)
  File "/usr/lib/python2.7/dist-packages/pip/commands/install.py", line 283, in run
    requirement_set.install(install_options, global_options, root=options.root_path)
  File "/usr/lib/python2.7/dist-packages/pip/req.py", line 1436, in install
    requirement.install(install_options, global_options, *args, **kwargs)
  File "/usr/lib/python2.7/dist-packages/pip/req.py", line 672, in install
    self.move_wheel_files(self.source_dir, root=root)
  File "/usr/lib/python2.7/dist-packages/pip/req.py", line 902, in move_wheel_files
    pycompile=self.pycompile,
  File "/usr/lib/python2.7/dist-packages/pip/wheel.py", line 206, in move_wheel_files
    clobber(source, lib_dir, True)
  File "/usr/lib/python2.7/dist-packages/pip/wheel.py", line 193, in clobber
    os.makedirs(destsubdir)
  File "/usr/lib/python2.7/os.py", line 157, in makedirs
    mkdir(name, mode)
OSError: [Errno 13] Permission denied: '/usr/local/lib/python2.7/dist-packages/amqp-1.4.6.dist-info'

Option a) Create a virtualenv, activate it and install:

virtualenv .venv
source .venv/bin/activate
pip install -r requirements.txt

Option b) Install in your homedir:

pip install --user -r requirements.txt

My recommendation use safe (a) option, so that requirements of this project do not interfere with other projects requirements.

We should really stop advising the use of sudo with pip install. It’s better to first try pip install --user. If this fails then take a look at the top post here.

The reason you shouldn’t use sudo is as follows:

When you run pip with sudo, you are running arbitrary Python code from the Internet as a root user, which is quite a big security risk. If someone puts up a malicious project on PyPI and you install it, you give an attacker root access to your machine.

You are trying to install a package on the system-wide path without having the permission to do so.

1. In general, you can use sudo to temporarily obtain superuser permissions at your responsibility in order to install the package on the system-wide path:

    sudo pip install -r requirements.txt
   

   Find more about sudo here.

   Actually, this is a bad idea and there’s no good use case for it, see @wim’s comment.

2. If you don’t want to make system-wide changes, you can install the package on your per-user path using the --user flag.

   All it takes is:

    pip install --user runloop requirements.txt
   

3. Finally, for even finer grained control, you can also use a virtualenv, which might be the superior solution for a development environment, especially if you are working on multiple projects and want to keep track of each one’s dependencies.

   After activating your virtualenv with

   $ my-virtualenv/bin/activate

   the following command will install the package inside the virtualenv (and not on the system-wide path):

   pip install -r requirements.txt

Just clarifying what worked for me after much pain in linux (ubuntu based) on permission denied errors, and leveraging from Bert’s answer above, I now use …

$ pip install --user <package-name>

or if running pip on a requirements file …

$ pip install --user -r requirements.txt

and these work reliably for every pip install including creating virtual environments.

However, the cleanest solution in my further experience has been to install python-virtualenv and virtualenvwrapper with sudo apt-get install at the system level.

Then, inside virtual environments, use pip install without the --user flag AND without sudo. Much cleaner, safer, and easier overall.

User doesn’t have write permission for some Python installation paths. You can give the permission by:

sudo chown -R $USER /absolute/path/to/directory

So you should give permission, then try to install it again, if you have new paths you should also give permission:

sudo chown -R $USER /usr/local/lib/python2.7/

If you need permissions, you cannot use ‘pip’ with ‘sudo’. You can do a trick, so that you can use ‘sudo’ and install package. Just place ‘sudo python -m …’ in front of your pip command.

sudo python -m pip install --user -r package_name

So, I got this same exact error for a completely different reason. Due to a totally separate, but known Homebrew + pip bug, I had followed this workaround listed on Google Cloud’s help docs, where you create a .pydistutils.cfg file in your home directory. This file has special config that you’re only supposed to use for your install of certain libraries. I should have removed that disutils.cfg file after installing the packages, but I forgot to do so. So the fix for me was actually just…

rm ~/.pydistutils.cfg.

And then everything worked as normal. Of course, if you have some config in that file for a real reason, then you won’t want to just straight rm that file. But in case anyone else did that workaround, and forgot to remove that file, this did the trick for me!

It is due permission problem,

sudo chown -R $USER /path to your python installed directory

default it would be /usr/local/lib/python2.7/

or try,

pip install --user -r package_name

and then say, pip install -r requirements.txt this will install inside your env

dont say, sudo pip install -r requirements.txt this is will install into arbitrary python path.