[DOCKER] How to easily create an environment where python code runs on Jupyter without polluting the local environment

Overview

--Create a state where ** python code can be run ** without polluting the local environment -** Time required 5 minutes **

--The article assumes that it will be built on a macbook, but if you want to start it on a server and access it with a browser, replace localhost with any server and execute it (and set a port that can be accessed by a browser).

Install Docker

https://docs.docker.com/docker-for-mac/install/#download-docker-for-mac

--Click Get Docker for mac to get Docker.dmg --Start Docker.dmg and install it on your mac

Start container with Anaconda docker image

Launch Terminal

Download the required docker image

Get your favorite docker image, here we will use the image that contains anaconda3 https://hub.docker.com/r/continuumio/anaconda3/~/dockerfile/

$ docker pull continuumio/anaconda3
Using default tag: latest
latest: Pulling from continuumio/anaconda3
8ad8b3f87b37: Pull complete 
e8b8d30f7444: Pull complete 
...
Status: Downloaded newer image for continuumio/anaconda3:latest

Create a working directory referenced by jupyter notebook

$ mkdir -p $HOME/jupyter-notebook/notebooks

Start Docker container

$ docker run \
    -d -i -t \
    -p 8888:8888 \
    --name jupyter-notebook \
    -v $HOME/jupyter-notebook/notebooks:/opt/notebooks \
    continuumio/anaconda3 \
    /bin/bash -c "/opt/conda/bin/conda install jupyter -y --quiet && /opt/conda/bin/jupyter notebook --notebook-dir=/opt/notebooks --ip='*' --no-browser --allow-root"

-p 8888: xxxx If the Jupyter Notebook default 8888 is not convenient, replace it with another port xxxx.

Checking the status of the container

$ docker ps -a
CONTAINER ID        IMAGE                   COMMAND                  CREATED             STATUS              PORTS                    NAMES
96f9690341ae        continuumio/anaconda3   "/usr/bin/tini -- ..."   6 seconds ago       Up 5 seconds        0.0.0.0:8888->8888/tcp   jupyter-notebook

Open JupyterNotebook from your browser

Get the URL of the Jupyter Notebook running on the container

In the log obtained by the docker logs command, the URL with the token for accessing JupyterNotebook is recorded, so get it.

$ docker logs jupyter-notebook

Fetching package metadata .........
Solving package specifications: .
...
    
    Copy/paste this URL into your browser when you connect for the first time,
    to login with a token:
        http://localhost:8888/?token=a5b5211197b4d9b2f8f93ee73c085786039486369c5bcbec <<<<---- here

Open the access URL in your browser

Open the obtained URL in a browser (if you have changed the port, replace it with the specified port)

http://localhost:8888/?token=a5b5211197b4d9b2f8f93ee73c085786039486369c5bcbec

Opened No code yet

Code execution

Place your favorite code in your working directory As an example, git clone Deep Learning repository from scratch.

$ cd $HOME/jupyter-notebook/notebooks

$ git clone https://github.com/oreilly-japan/deep-learning-from-scratch.git
Cloning into 'deep-learning-from-scratch'...
remote: Counting objects: 322, done.
remote: Total 322 (delta 0), reused 0 (delta 0), pack-reused 322
Receiving objects: 100% (322/322), 4.87 MiB | 944.00 KiB/s, done.
Resolving deltas: 100% (160/160), done.
Checking connectivity... done.

Immediately reflected on Jupyter Notebook

(The acquired file is a text file, so it cannot be executed as it is) Create an executable file such as Python3 from the New pull-down on the Jupyter screen, and cut and paste the code you want to execute to move the code freely.

that's all! : D

If you don't need a Docker container anymore

Let's destroy the container with docker stop, docker rm :)

$ docker ps -a 
CONTAINER ID        IMAGE                   COMMAND                  CREATED             STATUS              PORTS                    NAMES
96f9690341ae        continuumio/anaconda3   "/usr/bin/tini -- ..."   6 seconds ago       Up 5 seconds        0.0.0.0:8888->8888/tcp   jupyter-notebook

$ docker stop 96f9690341ae
$ docker rm 96f9690341ae