python_docker_for_ubuntu

Python data analysis on Ubuntu

A Docker image for Python data analysis on Ubuntu 16.04. Installed Tools:

Pyenv
Python 3.5.2
Pandas
Scipy
Numpy
Scikit-learn
Matplotlib
Seaborn
Jupyter-notebook
Beautiful soup 4

Setup instructions

I ubuntu installed in virtual box for running the Jupyter notebook sessions. The setup instructions can be find below

• Install docker and docker.io
• Create docker group

sudo groupadd docker

• Add user to the group docker sudo usermod -aG docker $USER

• Start docker daemon

sudo service docker restart

For ubuntu > 14.04

sudo service docker.io restart  

• Pull docker image

docker pull avikdatta/python_data_docker_files

• Check the available images

docker images

• Run Jupyter session

docker run -v /home/$user/data_dir:/home/vmuser/data \
           -p 8888:8888                              \
           --net=host                                \
           avikdatta/python_data_docker_files:latest \
           jupyter-notebook  --ip 0.0.0.0

• Access the Jupyter secure session using the IP of the server/ virtual machine. You will need the token for accessing this instance

Installing new packages

Any new packages/tools can be installed in the docker image using the interactive mode

• Run bash from docker

docker run -v /home/$user/data_dir:/home/vmuser/data \
           -it avikdatta/python_data_docker_files:latest \
           /bin/bash

• Install any tools using apt-get

• Detach the docker container, use Control+pq

• Get the container id

docker ps

• Commit changes using a new tag name

docker commit $container_id avikdatta/python_data_docker_files:new_feature

• Kill container

docker kill $container_id

• Check the available images

docker images

• Run new image

docker run -v /home/$user/data_dir:/home/vmuser/data      \
           -p 8888:8888                                   \
           --net=host                                     \
           avikdatta/python_data_docker_files:new_feature \
           jupyter-notebook --ip 0.0.0.0