Visual Studio for Data Science

October 3, 2018

Should you use VS for Data Science?

 

Actually, no – you shouldn’t use Visual Studio if you are doing exploratory work, but if you are shipping data products into a production system then yes you really should!

 

If you are doing exploratory work or investigative work there are better IDEs out there for you to use. R Studio or Spyder will do nicely.

 

If you would like to document your code, graphs and text in one place for reproducibility, team learning etc, well use Jupter labs or Jupyter notebooks or R Markdown.

 

 

What are the benefits of Visual Studio for Data Science?

  • Visual Studio is an IDE for the developer, data scientists are not developers, but in some roles the line between dev and data scientist is quite thin.

  • There are amazing templates for machine learning, Flask apps etc to get you started in your projects

  • Better for executing MS team data science process – MS Team Services and Git integration

  • Great support and integration for SQL Server and other databases

  • Just better debugging, unit tests, Intellisense, highlighting, code flows off your fingers

 

 

What are the disadvantages

  • Load times, can be clunky. Sometimes it hangs, the best option is just to restart.

  • The learning curve is pretty dang high if you are a “sometimes coder”

  • It certainly does chew up some memory

 

 

How do I get started with Visual Studio?

 

1) Download Visual Studio Community Edition

 

 

https://visualstudio.microsoft.com/downloads/

 

 

2) Select the “Data Science and analytical applications checkbox during the install process”

 

 

 

Get really comfortable, because this is going to take ages and ages.

 

 

 

3) Restart your computer and then launch Visual Studio 2017. Curse Microsoft a little bit.

 

 

4) Launch Visual Studio and sign in using your Microsoft Account. Get a Microsoft Account here.

 

https://account.microsoft.com

 

 

5) Boom! We are go!

 

 

 

Now Check out what we have when we go to File > New Project. We have template code for classifications, regressions and clustering projects!

 

 

 

Want to know how to build a Flask app in Python in about 5 seconds? File > Web > Flask Web Project

 

 

 

There are extra packages required, and the default option is to have them installed in a virtual environment. This is a great option, in fact managing your virtual environment with right click installs is just a beautiful thing!

 

Boom! Check that out! We have template code for our Flask App all good to go!

 

 

 

Then we run the app. There we go! A Flask App in 5 mins!

 

 

 

More resources below:

 

Python Tools for Visual Studio

 

https://www.youtube.com/watch?v=_okUV47eM5c

 

R Tools for Visual Studio

 

https://www.youtube.com/watch?v=dll3IS1bfWQ

 

 

Getting started with the Visual Studio IDE

https://mva.microsoft.com/en-US/training-courses/getting-started-with-visual-studio-2017-17798?l=9oIw0FD6D_3611787171

 

 

 

Please reload

Recent Posts

Please reload

Archive

Please reload

Tags

  • Black Facebook Icon

©2018 BY DATAFRIENDS.