Develop Anaconda applications in Python3. Installs dependencies from your environment.yml file and the Python extension.

See history for information on the contents of published images.

You can directly reference pre-built versions of .devcontainer/Dockerfile by using the image property in .devcontainer/devcontainer.json or updating the FROM statement in your own Dockerfile to the following. An example Dockerfile is included in this repository.

• mcr.microsoft.com/devcontainers/anaconda

Refer to this guide for more details.

You can decide how often you want updates by referencing a semantic version of each image. For example:

• mcr.microsoft.com/devcontainers/anaconda:1-3
• mcr.microsoft.com/devcontainers/anaconda:1.3-3
• mcr.microsoft.com/devcontainers/anaconda:1.3.11-3

See history for information on the contents of each version and here for a complete list of available tags.

Alternatively, you can use the contents of .devcontainer to fully customize your container's contents or to build it for a container host architecture not supported by the image.

This dev container and its associated image includes the conda package manager. Additional packages installed using Conda will be downloaded from Anaconda or another repository if you configure one. To reconfigure Conda in this container to access an alternative repository, please see information on configuring Conda channels here.

Access to the Anaconda repository is covered by the Anaconda Terms of Service, which may require some organizations to obtain a commercial license from Anaconda. However, when this dev container or its associated image is used with GitHub Codespaces or GitHub Actions, all users are permitted to use the Anaconda Repository through the service, including organizations normally required by Anaconda to obtain a paid license for commercial activities. Note that third-party packages may be licensed by their publishers in ways that impact your intellectual property, and are used at your own risk.

By default, frameworks like Flask only listens to localhost inside the container. As a result, we recommend using the forwardPorts property (available in v0.98.0+) to make these ports available locally.

"forwardPorts": [5000]

The appPort property publishes rather than forwards the port, so applications need to listen to * or 0.0.0.0 for the application to be accessible externally. This conflicts with the defaults of some Python frameworks, but fortunately the forwardPorts property does not have this limitation.

Given JavaScript front-end web client code written for use in conjunction with a Python back-end often requires the use of Node.js-based utilities to build, you can use a Node feature to install any version of Node by adding the following to devcontainer.json:

{
  "features": {
    "ghcr.io/devcontainers/features/node:1": {
      "version": "latest"
    }
  }
}

This image is based on the ContinuumIO/anaconda3 docker image, which has all the anaconda3 packages from defaults installed into the base conda's environment. It is recommended not to install packages from different channels in one environment since it could cause conflicts. When installing a package from a different channel (e.g., conda-forge) is required, the better approach is to create a new conda environment.

conda create --name <env_name> -c <channel> --yes <package_name>

As covered in the user FAQ for Anaconda, you can install different versions of Python than the one in this image by running the following from a terminal:

conda install python=3.6

Or in a Dockerfile:

RUN conda install -y python=3.6

For convenience, this image will automatically install dependencies from the environment.yml file in the parent folder when the container is built. You can change this behavior by altering this line in the .devcontainer/Dockerfile:

RUN if [ -f "/tmp/conda-tmp/environment.yml" ]; then /opt/conda/bin/conda env update -n base -f /tmp/conda-tmp/environment.yml; fi \
    && rm -rf /tmp/conda-tmp

Use this container to run Jupyter notebooks.

1. Edit the ./.devcontainer/devcontainer.json file and add 8888 in the forwardPorts array:

   // Use 'forwardPorts' to make a list of ports inside the container available locally.
   "forwardPorts": [8888],

.

1. Edit the ./.devcontainer/devcontainer.json file and add a postStartCommand command to start the Jupyter notebook web app after the container is created. Use nohup so it isn't killed when the command finishes. Logs will appear in nohup.out.

   // Use 'postStartCommand' to run commands after the container is created.
   "postStartCommand": "nohup bash -c 'jupyter notebook --ip=0.0.0.0 --port=8888 --allow-root &'",

2. View the terminal output to see the correct URL including the access token:

    http://127.0.0.1:8888/?token=1234567

3. Open the URL in a browser. You can edit and run code from the web browser.

Copyright (c) Microsoft Corporation. All rights reserved.

Licensed under the MIT License. See LICENSE