Installing FiftyOne¶

This page describes how to install FiftyOne in a virtual environment.

Using a virtual environment is strongly recommended because it allows maintaining an isolated environment in which FiftyOne and its dependencies can be installed. FiftyOne has many dependencies, some versions of which may conflict with versions already installed on your machine.

Prerequisites¶

You will need a working Python installation. FiftyOne currently requires Python 3.6 or newer. On Linux, we recommended installing Python through your system package manager (APT, YUM, etc.) if it is available. On other platforms, Python can be downloaded from python.org. To verify that a suitable Python version is installed and accessible, run python3 --version or python --version.

Creating a virtual environment¶

First, identify a suitable Python executable. On many systems, this will be python3 , but it may be python on other systems instead. To confirm your Python version, pass --version to Python. Here is example output from running these commands:

$ python --version
Python 2.7.17
$ python3 --version
Python 3.6.9

In this case, python3 should be used in the next step.

Navigate to a folder where you would like to create the virtual environment. Using the suitable Python version you have identified, run the following to create a virtual environment called env (you can choose any name):

python3 -m venv env

Replace python3 at the beginning of a command if your Python executable has a different name. This will create a new virtual environment in the env folder, with standalone copies of Python and pip, as well as an isolated location to install packages to. However, this environment will not be used until it is activated. To activate the virtual environment, run the following command:

. env/bin/activate

Be sure to include the leading .

env\Scripts\activate.bat

After running this command, your shell prompt should begin with (env) , which indicates that the virtual environment has been activated. This state will only affect your current shell, so if you start a new shell, you will need to activate the virtual environment again to use it. When the virtual environment is active, python without any suffix will refer to the Python version you used to create the virtual environment, so you can use this for the remainder of this guide. For example:

$ python --version
Python 3.8.3

Also note that python and pip live inside the env folder (in this output, the path to the current folder is replaced with ...):

$ which python
.../env/bin/python
$ which pip
.../env/bin/pip

> where python
...\env\Scripts\python.exe
C:\Program Files\Python38\python.exe
> where pip
...\env\Scripts\pip.exe
C:\Program Files\Python38\Scripts\pip.exe

Before you continue, you should upgrade pip and some related packages in the virtual environment. FiftyOne’s packages rely on some newer pip features, so older pip versions may fail to locate a downloadable version of FiftyOne entirely. To upgrade, run the following command:

pip install --upgrade pip setuptools wheel

More virtual environment resources¶

If you ever want to leave an activated virtual environment and return to using your system-wide Python installation, run deactivate.

There are lots of ways to set up and work with virtual environments, some of which are listed here. These may be particularly useful to review if you are dealing with virtual environments frequently:

The venv module used in this guide is documented here, with information on additional arguments that the venv command accepts.
There is a similar virtualenv package (pip install virtualenv) that supports older Python versions.
virtualenvwrapper adds some convenient shell support for creating and managing virtual environments.

Installing FiftyOne¶

Note

FiftyOne is currently in private beta! If you have registered, your welcome email contains a token that you will need to run some of the commands below. Replace all instances of YOUR_TOKEN below with your token.

If you haven’t registered for the FiftyOne Beta, you can sign up at https://voxel51.com/fiftyone/#beta

To install FiftyOne in a virtual environment, ensure that the virtual environment is active as described in the previous section, and then run the command below:

pip install --index https://YOUR_TOKEN@pypi.voxel51.com fiftyone

This will install FiftyOne and all of its dependencies, which may take some time. Once this has completed, you can verify that FiftyOne is installed in your virtual environment:

$ python
Python 3.6.9 (default, Apr 18 2020, 01:56:04)
[GCC 8.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>>
>>> import fiftyone as fo
>>> fo.__file__
'.../env/lib/python3.X/site-packages/fiftyone/__init__.py'
>>>
>>> session = fo.launch_app()
>>> exit()

Linux users: if you encounter an error related to MongoDB failing to start, such as “Could not find mongod”, you may need to install additional packages. See the troubleshooting section for details.

Windows users: If you encounter errors related to missing msvcp140.dll, you will need to install the 64-bit Visual Studio 2015 C++ redistributable library, available here (choose the x64 version).

Installing extra packages¶

Various tutorials and guides that we provide on this site require additional packages in order to run. If you encounter a missing package, you will see helpful instructions on what you need to install. Or, you can preemptively install what you’ll need by installing the following additional packages via pip in your virtual environment:

ipython to follow along with interactive examples more easily
tensorflow for examples requiring TensorFlow. The installation process can vary depending on your system, so consult the Tensorflow documentation for specific instructions.
tensorflow-datasets for examples that rely on loading TensorFlow datasets
torch and torchvision for examples requiring PyTorch. The installation process can vary depending on your system, so consult the PyTorch documentation for specific instructions.

For your own work, FiftyOne does not strictly require any of these packages, so you can install only what you need.

Upgrading FiftyOne¶

Passing the --upgrade (or -U) option to pip install can be used to upgrade an existing FiftyOne installation:

pip install --index https://YOUR_TOKEN@pypi.voxel51.com --upgrade fiftyone

Uninstalling FiftyOne¶

FiftyOne and all of its subpackages can be uninstalled with:

pip uninstall fiftyone fiftyone-brain fiftyone-db fiftyone-gui

Troubleshooting¶

Installing MongoDB on Linux¶

FiftyOne relies on a version of MongoDB that works on Ubuntu 18.04 and several other modern distributions. If this version does not work on your distribution, there are alternative builds available, or you can use an existing installation of MongoDB.

Alternative builds¶

Alternative builds are available as pip packages for the distributions listed below, and can be installed by running the corresponding command. Note that these packages must be installed after the fiftyone package; if you install fiftyone afterwards, you can fix your MongoDB installation by adding --force-reinstall to the commands below.

pip install --index https://YOUR_TOKEN@pypi.voxel51.com fiftyone-db-ubuntu1604

pip install --index https://YOUR_TOKEN@pypi.voxel51.com fiftyone-db-debian9

Manual installation¶

FiftyOne also supports using an existing MongoDB installation (version 3.6 or newer). This can be installed through many distributions’ package managers. Note that only the mongod (server) binary is required, so you may not need the complete MongoDB package. For example, Debian-based distributions make this available in the mongodb-server package.

If your distribution does not provide a new-enough version of MongoDB, or if you would like to install a newer version, see the MongoDB documentation for instructions on installing MongoDB on your distribution. Note that you only need the mongodb-org-server package in this case.

To verify the version of your MongoDB installation, run mongod --version, which should produce output that looks like this:

db version v4.2.6
git version: 20364840b8f1af16917e4c23c1b5f5efd8b352f8
OpenSSL version: OpenSSL 1.1.1  11 Sep 2018
allocator: tcmalloc
modules: none
build environment:
    distmod: ubuntu1804
    distarch: x86_64
    target_arch: x86_64

Verify that the version after “db version” is at least 3.6.