Install Guide

This guide is a resource for using the Allen SDK package. It is maintained by the Allen Institute for Brain Science.


As of October 2019, we have dropped Python 2 support. The Allen SDK is developed and tested with Python 3.6 and 3.7. We do not guarantee consistent behavior with other Python versions.

Quick Start Using Anaconda

  1. From the Anaconda downloads page, download the Python 3.7 version for your operating system and run the installer.

  2. After the installation is complete, open up a terminal (in Windows open Anaconda3 Command Prompt).

  3. Install the AllenSDK using PIP:

    pip install allensdk
  4. Download one of our many Jupyter Notebook examples to a new folder.

  5. In your terminal, navigate to the directory where you downloaded the Jupyter Notebook example and run the following command:

    jupyter notebook
  6. Your browser should open and you should see the Jupyter Notebook example. Enjoy using the Allen SDK!

Quick Start Using Pip

First ensure you have pip installed. It is included with the Anaconda distribution.

pip install allensdk

To uninstall the SDK:

pip uninstall allensdk

Other Distribution Formats

The Allen SDK is also available from the Github source repository.

Installation with Docker (Optional)

Docker is an open-source technology for building and deploying applications with a consistent environment including required dependencies. The AllenSDK is not distributed as a Docker image, but example Dockerfiles are available.

  1. Ensure you have Docker installed.

  2. Use Docker to build the image:

    docker pull alleninstitute/allensdk

    Other docker configurations are also available under docker directory in the source repository.

  3. Run the docker image:

    docker run -i -t -p 8888:8888 alleninstitute/allensdk /bin/bash
  4. Run the SDK tests:

    cd allensdk
    make test
  5. Start a Jupyter Notebook:

    cd allensdk/doc_template/examples_root/examples/nb
    jupyter notebook --ip=* --no-browser --allow-root

    Using the browser on your host machine, navigate to the path provided by the output from the jupyter notebook command.