Python Notebook with Pandas in One Line

If you have Docker installed, opening up a Jupyter Python notebook with Pandas for data analysis can be done with one step on the command line. Running python notebooks in a Docker container prevents tons of headaches with installation, permissions, version conflicts, etc.

Docker Command

docker run --name notebookContainer --rm -p 127.0.0.1:8888:8888 -v pwd:/home/jovyan/work jeremyworks/jupyter-scipy:1.0.0

The first time you run this command it will cache all the images to run it quickly next time.

Docker Command Explained

  • docker run – create and start a new container
  • –name notebookContainer – name the new container to make it easier to identify
  • –rm – remove the container when the docker run command finishes
  • -p 127.0.0.1: 8888:8888 – map port 8888 of the container to port 8888 of localhost(127.0.0.1)
  • -v `pwd`:/home/jovyan/work – map the current working directory to the notebook working directory. By mapping this volume to the same directory you can save and retrieve notebooks created during previous Docker run commands
  • jeremyworks/jupyter-scipydocker – docker image to run. When the version tag is left out this is the same as specifying jeremyworks/jupyter-scipydocker:latest

Browser Access

Paste the URL copied from the console output starting with, http://127.0.0.1:8888/ , into a browser.

This URL includes a token for access to the new Jupyter server running in Docker.

Create a New Notebook

Select New/ Python 3

Try it out

Here’s some example code to create a DataFrame

import pandas as pd
import numpy as np

df = pd.DataFrame(np.random.randint(0,100,size=(10, 3)), columns=['Col_A','Col_B','Col_C'])
df

Links

Leave a Comment

Your email address will not be published. Required fields are marked *