Tried 100 Knocks by Japan Data Scientist Society

Japan Data Science Society published a set of 100 questions, answers and programming environments for people who wanted to learn data science on GitHub. This package is provided in a form of Docker. In the Docker container, Python, R and SQL are all set up, so you can write a code and run the code freely. 

I have never used Docker before, so I thought this was a good opportunity for me to learn Docker.

To be honest, the setup was not the easiest thing for a beginner like me. I spent about a good full day until I start using it. Unfortunately, the contents are written in only Japanese but I would like to share the experience in setting up in the case they may provide English version in the future.

Prerequisite

You need to download and install:
1. Git from git-scm.com
This is the only screen you really have to pay attention because if you accept the default first option, the core.autocrlf value will be set to 'true' and SQL database cannot be built properly.

DO NOT forget to select the 2nd option!

2. Docker Desktop for Windows from Hub.Docker.com 

Once these software is installed, open PowerShell and enter the commands below one by one.

Open PowerShell and run the below commends one by one

Create Docker Container

1. Git clone https://github.com/The-Japan-DataScientist-Society/100knocks-preprocess.git

https://github.com/The-Japan-DataScientist-Society/100knocks-preprocess.git is a 100 knock training material download URL. Make sure the link is correct before running the command, The above link is active as of the end of November 2020

    

    Done!

2. cd 100knocks-preprocess

Command #1 creates a copy of the 100Knocks-preprocess folder right underneath your profile.

Move to the folder. 

3. Dcocker-compose up -d --build

This process took about 10 minutes with my new Dell Latitude 9510 (Win 10 pro 2004 X64, Intel Core i5, 16 GB RAM)

 

    Done!

Access Docker

Open the browser and type http:// localhost:8888.

Jupyter Notebook will open.

To end Docker

docker-compose stop

docker-machine stop default

To resume Docker

cd 100knocks-preprocess

docker-compose start