Data Exploration and Predictive Modeling (Tutorial: SQL Server R Services)


Updated: December 4, 2015

MicrosoftSQL Server 2016 (CTP3) contains a preview version of R Services (In-database). This new product includes provides closer integration of R with enterprise data sources such as SQL Server databases.

In This Topic

Product samples for SQL Server 2016 are available from the Microsoft Download Center:

SQL Server 2016 Product Samples

This download includes the AdventureWorks databases, as well as samples for new features such as Stretch Database.

System_CAPS_ICON_tip.jpg Tip

To get only the samples for R Services (In-database), select the zip file, and open the folder Advanced Analytics.

Developing an End-to-End Data Science Solution

Data Science End-to-End Walkthrough

This solution uses the New York City Taxi data set to demonstrate the entire data science process:

  • Import data to SQL Server

  • Build a predictive model using R

  • Deploy the R model to SQL Server

  • Use the model in production for scoring

Start here if you want to learn how to use the new functions provided by the RevoScaleR package to work with SQL Server data in R.

System_CAPS_ICON_note.jpg Note

For additional examples of how the New York City taxi dataset is used in machine learning, see Building Azure ML Models on the NYC Taxi Dataset.

SQL Server R Services for the Data Scientist

The following walkthrough, currently available in PDF format from the Microsoft Download Center, assumes that you are already familiar with the R language and want to learn how to use the new database integration features, such as the ability to change execution context.

Getting Started with RevoScaleR and SQL Server

The walkthrough explains the functionality of the new R packages and functions provided by Revolution R Enterprise 7.5.0, including:

  • Defining a SQL Server data source

  • Uploading data from a CSV file into a SQL Server database by using R

  • Using R to modify tables and add column metadata

  • Using the enhanced RevoScale R packages in SQL Server

  • Creating predictive models and storing them in SQL Server

  • Using stored models to generate predictions

Depending on your familiarity with R, the walkthrough should take between 60 and 90 minutes to complete.

This blog post outlines the process of using the R packages and functions provided by R Services (In-database) to connect to SQL Server. Sample code is provided.

Using R inside SQL Server

System_CAPS_ICON_note.jpg Note

If you need more information about learning the R language, or want to learn more about the R packages provided in R Services (In-database), see these resources.

Getting Started with SQL Server R Services
SQL Server R Services Features and Tasks