Installation¶

You can install Atoti and its JupyterLab extension with conda (recommended) or pip.

With Conda¶

1. Install Conda¶

Important

Make sure to use a 64-bit version since some packages are not available with Conda 32-bit like openjdk’s recent version.

Install Miniconda 64-bit (recommended) or Anaconda 64-bit.

2. Set up conda-forge channel¶

Add conda-forge to your Conda channels:

conda config --add channels conda-forge

3. Create a new Conda environment¶

Create a new Conda environment and activate it:

conda create --name atoti
conda activate atoti

4. Install Atoti¶

The easiest way to get started is to use the single-package Atoti bundle to install the library, its JupyterLab extension, and all their dependencies:

conda install atoti-bundle

Alternatively, if you are comfortable with Conda and would like a more fine-grained control on the packages being installed in your environments, you can install these packages instead:

conda install atoti jupyterlab-atoti jupyterlab=2 openjdk=11 python=3.7

5. Rebuild JupyterLab¶

JupyterLab needs to be rebuilt to include the Atoti extension. You can either wait until you open your first JupyterLab session and click on the Build button when prompted, or pro-actively run the following command for optimal frontend performance (it can take several minutes):

Windows¶

set NODE_OPTIONS=--max-old-space-size=4096 & jupyter lab build

Linux / macOS¶

NODE_OPTIONS=--max-old-space-size=4096 jupyter lab build

Advanced¶

Atoti+¶

Atoti+ is our entreprise edition, it requires a license.

First, follow steps 1. Install Conda and 2. Set up conda-forge channel.

With Artifactory¶

We distribute our enterprise packages through Artifactory. You can add our Conda channel to your Conda configuration:

Open this page and fill the authentication form with your ActiveViam credentials to get your encrypted password. If you don’t have any credentials, please contact us at support@atoti.io.

Add the ActiveViam channel to your Conda channels, replacing {{username}} with your ActiveViam username and {{encrypted_password}} with the encrypted password you obtained previously:

conda config --add channels https://{{username}}:{{encrypted_password}}@activeviam.jfrog.io/activeviam/api/conda/conda-internal

Install our bundle:

conda install atoti-plus-bundle

Or our separate packages:

conda install atoti-plus jupyterlab-atoti jupyterlab=2 openjdk=11 python=3.7

Finally, follow step 5. Rebuild JupyterLab.

Without Artifactory¶

If your network policy is blocking our Artifactory repository, you will have to use this less convenient setup and upgrades won’t be streamlined.

First, you need to have the artifacts of our Conda packages on your computer. Let’s say you have our bundle at /home/me/project/atoti-bundle-0.2.0-0.tar.bz2. Then, to install it, run these commands:

# Create a local Conda channel first.
conda install conda-build
mkdir -p /home/me/project/tmp/channel/noarch
mv /home/me/project/atoti-bundle-0.2.0-0.tar.bz2 /home/me/project/tmp/channel/noarch
conda index /home/me/project/tmp/channel/
# Do the actual installation.
conda install --channel file://home/me/project/tmp/channel atoti-bundle

Finally, follow step 5. Rebuild JupyterLab.

Without Conda¶

If you would rather use pip/Pipenv/Poetry than Conda, the library is also available as a Python Wheel and the JupyterLab extension as an npm package. Contact us for more information at support@atoti.io.