Hands-On Data Science with R

Dr Graham Williams, PhD (ANU, Machine Learning), BSc (Maths, Hons)
Data Scientist, Togaware and Australian Taxation Office
Adjunct Professor, Australian National University and University of Canberra
International Visiting Professor, Chinese Academy of Sciences

We begin with an overview of how an organization should go about setting up their Analytics capability and then introduce the Data Scientist to the most fully featured yet cost effective toolkit available: R.

But be aware, ICT departments will see Analytics like they see Accounting software. They will survey the available products (and today there's hundreds of products available), seek advise and understanding from the vendors (rather than from the practitioners), decide on one provider (the one true solution), purchase that product and required infrastructure, and believe that they have delivered an Analytics capability to support the organization for the next 10 years, just like they would for an Accounting capability. I've seen it, I've seen millions wasted on software, and I've seen ICT departments simply not getting it. Analytics is about the skill of a Data Scientist, using a variety of ever changing tools on platforms that are also changing quickly. Instead of expensive deployments that will live stably for the next 10 years, be prepared for inexpensive initial deployments that will grow and change rapidly, and some of which might fail quickly. See Analyst First for further views along these lines.

This web site provides extensive material for the Data Scientist. Togaware also provides a unique offering of in-situ hands-on training. We offer traditional out-of-office training courses, but we find more effective learning can occur hands-on in-situ. We offer one of the world's leading Data Scientists to work alongside and mentor your staff over one or two weeks. We work confidentially on actual projects, with training "on-the-job" provided by a professional with 30 years experience in the industry and author of the best selling book on Data Mining with Rattle and R. Contact Togaware Training at training@togaware.com for details.

Our on-line resources, including Hands-On Data Science, weave together a collection of freely available and open source tools for the Data Scientist. The tools are all part of the R Statistical Software Suite. Each chapter is made up of multiple pages, but each page within a chapter is a one page guide that covers a particular aspect of the topic (hence also refered to as the OnePageR guide). They are a great place to start, before engaging our hands-on training experts.

Hands-On Data Science can be worked through as a hands-on guide and then used as a reference guide. Each page aims to be a bite sized chunk for hands-on learning, building on what has gone before. Many chapters also have a lecture pack and a laboratory session where a number of tasks can be completed. The R code sitting behind each chapter is also provided and can be easily run standalone to replicate the material presented in the chapter.

The material is always under development! Chapters will change (and hopefully improve) regularly. Links preceded with a * are more well developed. All of the material is provided under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License allowing access to everyone for any purpose (except commercial) and is provided at no cost. You can assist in helping cover the costs of providing this material through a $40 contribution using PayPal. Your support encourages further development of this resource as does feedback, suggestions, and ideas, which are always welcome.

Refer to the Data Mining Survival Guide or my book on Data Mining with Rattle and R: The Art of Excavating Data for Knowledge Discovery (Use R) for related material.

Many of the initial chapters were developed and tested whilst visiting the Shenzhen Institutes of Advanced Technology as an International Visiting Professor of the Chinese Academy of Sciences.

The data used across the chapters is available for download as data.zip.


Part 1: Getting Started as a Data Scientist
  1. Data Scientists and Analytics *Lecture - *Chapter - *R
  2. Rattle to R: *Chapter - *R
Part 2: R for the Eager Data Scientist
  1. A Template for Preparing Data: *Chapter - *R
  2. A Template for Building Models: *Chapter - *R
  3. Case Studies: *Chapter - *R
  4. The Basics of R Chapter - R
Part 3: Dealing With Data
  1. Reading Data into R: *Chapter - *R
  2. Exploring and Summarising Data: *Chapter - *R
  3. Visualising Data with GGPlot2: *Chapter - *R
  4. Transforming Data: *Chapter - *R
Part 4: Descriptive Analytics
  1. Cluster Analysis: *Lecture - Chapter - R
  2. Association Analysis: *Lecture - Chapter - R
Part 5: Predictive Analytics
  1. Decision Trees: *Lecture - *Chapter - *R - *Rattle
  2. Ensembles of Decision Trees: *Lecture - *Chapter - *R
  3. Support Vector Machines
  4. Neural Networks
  5. Naive Bayes: Chapter - R
  6. Multivariate Adaptive Regression Splines: Chapter - R
  7. Evaluating Models: *Chapter - *R
  8. Scoring (R)
  9. PMML (R) Exporting Models for Deployment
Part 6: Advanced Analytics
  1. Text Mining: *Chapter - *R - Corpus as tar.gz or zip
  2. Social Network Analysis: Chapter - R
  3. Genetic Programming: Chapter - R
Part 7: Advanced R
  1. Strings: Chapter, R
  2. Dates and Time: *Chapter - *R
  3. Spatial Data *Chapter - *R
  4. Big Data *Chapter - *R
  5. Exploring Different Plots: Chapter - R
  6. Writing Functions: Chapter - R
  7. Parallel Processing: Chapter - R
  8. Environments: *Chapter - R
Part 8: Expert R
  1. Packaging (R) Pulling it Together into a Package
  2. Doing R with Style: *Chapter - *R
  3. Literate Data Science with KnitR: *Lecture - *Chapter - *R

Other great resources for modular approaches to learning R include:

Other Togaware resources:

Local package archive:

Creative Commons License

Shop at Amazon