nllab-Dockerfiles

Latest images on DockerHub:

nadanai263/nllab-python:003 - Python 3.8
nadanai263/nllab-jupyter:007 - Python 3.11.4, Julia 1.5.3
nadanai263/nllab-julia:005 - Julia 1.5

The numeric tags will increase as the containers are updated, so please check them on DockerHub.

Running Linux in a Docker container

As an example to start an interactive Linux shell which contains a Python environment, first navigate to the directory containing your work. Then (making sure you have a working Docker installation in place), run (on Mac, Linux)

docker run -it --rm -v "$PWD":/app nadanai263/nllab-python:003 /bin/bash

On Windows, replace "$PWD" with "%CD%" (for command prompt) or ${pwd} (for powershell).

This will pull and launch a Docker container, and start up a Linux shell. Your current working folder will be mounted to /app. Run your scripts as required.

Running Jupyter notebook in a Docker container

You can directly start a Jupyter notebook in your current directory, using

docker run -p 8888:8888 --rm -it -v "$PWD":/home/jovyan nadanai263/nllab-jupyter:007

Again, on Windows replace "$PWD" with "%CD%" (for command prompt) or ${pwd} (for powershell).

---

Useful Docker commands

1. Building and pushing to registry

Building from a Dockerfile: navigate to the directory containing the file, which must be named Dockerfile, and run

docker build -t newrepo . # Create image using this directory's Dockerfile and name it 'newrepo'

To upload our image 'newrepo', first login to Dockerhub, then tag:

docker login
docker tag newrepo username/repository:tag

where repository:tag is the new repository name and version, e.g. newrepo:001, and username is your Dockerhub username. Finally, push:

docker push username/repository:tag

2. Running

An example of a run command in more detail:

docker run -p 8888:8888 --rm -it -v "$PWD":/home/jovyan/ username/repository:tag

This command creates a Docker container from the online image, and launches it with the following options:

• -p 8888:8888 exposes the container's internal ports
• --rm deletes the container and associated files at exit
• -it runs in interactive mode
• -v "$PWD":/home/jovyan/ starts the container in your current directory
• username/repository:tag is the image:tag, corresponding to a specific image on DockerHub

3. More examples of run commands

Mac/Linux versions given; as always on Windows, replace "$PWD" with "%CD%" (for command prompt) or ${pwd} (for powershell).

To start a Linux shell interactively, mounting your current working directory to /app:

docker run -it --rm -v "$PWD":/app nadanai263/nllab-python:003 /bin/bash

To start a Jupyter notebook:

docker run -p 8888:8888 --rm -it -v "$PWD":/home/jovyan nadanai263/nllab-jupyter:007

To run a single python script called script.py located in your current directory:

docker run --rm -v "$PWD":/app nadanai263/nllab-python:003 python app/script.py

4. Maintenance and cleaning

To list all your local containers and images:

docker container ls -a
docker image ls -a

To remove a container, first run:

docker container rm <containerID>

Then remove image:

docker image rm <imageID>