JJ Allaire

J.J. Allaire is the Founder of RStudio and the creator of the RStudio IDE. J.J. is also the author of a wide variety of R packages including reticulate, htmlwidgets, rmarkdown, flexdashboard, distill, dygraphs, keras, and tensorflow.

Open Source Software for Data Science

Open-source software is fundamentally necessary to ensure that the tools of data science are broadly accessible, and to provide a reliable and trustworthy foundation for reproducible research. This talk will delve into why open source software is so important and discuss the role of corporations as stewards of open source software. I’ll also talk about how RStudio is structured and organized to pursue its mission of creating open source software for data science.

Jacquie Tran

Jacquie Tran is a sport scientist from Melbourne, Australia. She currently works as Team Leader for the Knowledge Edge for Tokyo team at High Performance Sport New Zealand. In this team, Jacquie uses qualitative and quantitative research methods to explore the aspects of preparation that contribute to successful performances at pinnacle events.

Data practices and data problems in performance sport

Sports performance data is abundant - with technological improvements, location tracking, video, inertial sensors, physiological monitoring, and activity logs have become commonplace in elite sport settings. In this talk, I’ll share some insights into how data is used in performance sport today and what opportunities lie ahead for sports organisations and professionals to improve their use of data.

Harkanwal Singh

Harkanwal is a data visualisation programmer, currently working at Westpac NZ. Previously, he established and led the data journalism desk at the New Zealand Herald. He is passionate about visualisation as code. He programs in R, JavaScript and Python. He enjoys teaching data visualisation in his own time. He loves reading words and code, both preferably of others.

The craft of data storytelling

Beyond the software, what are the elements of telling a compelling data story. The talk focuses on the importance of context and why you should never buy into the storytelling aspect beyond a certain point. Don’t fall in love with your stories, stay true to the data.

Anna Fergusson

Anna Fergusson is obsessed with teaching, in particular with developing inclusive, engaging, accessible, effective and fun ways to introduce people to learning statistics and data science. Before joining the University of Auckland as a Professional Teaching Fellow (Statistics), she was Head of Mathematics and Statistics at Avondale College for five years and a classroom teacher for 12 years. Anna is on the steering group the New Zealand Statistical Association education committee and is involved with the international collaboration between the New Zealand Statistical Association and the American Statistical Association. Her research has focused on exploring task and tool design for teaching statistical modelling. As part of her PhD research project, Anna is currently working closely with high school teachers to develop tools, tasks, assessment and learning programmes to support data science education at high school level.

Five ways to make learnr even better for teaching

The learnr package allows you to turn any R Markdown document into an interactive HTML-based tutorial with videos, code exercises, quiz questions and other interactive ‘shiny’ things. It’s a great tool to use for teaching R, and an even better tool when you exploit its potential to meet the needs of your learners. In this talk, I will share five ways that I have used learnr to engage introductory-level students with data science and discuss how each of these approaches enhanced the learning experience.

Simon Urbanek

Simon Urbanek is a Senior Lecturer of Data Science at the University of Auckland. He has been working at AT&T Labs in Data Science and AI Research for 15 years and is a member of the R Core Development Team with interests in visualization, interactive graphics, distributed computing and Big Data analytics. He is the author of RCloud and numerous R packages including Rserve, rJava, multicore, iotools, iPlots, RJDBC, Cairo and many others. Simon received his Ph.D. in Statistics at the University of Augsburg in 2004.

Scalable Collaborative Data Science using R

Data science encompasses the full analytic workflow starting with obtaining data, cleaning it, iteration over problem formulation, visualization, data modeling all the way to application and presentation. In real-world applications such process involves several collaborators, diverse tools and the scale of data can be in the hundreds of billions of data records. In this talk we will illustrate how R can be used by many users in a collaborative open-source RCloud environment to collaborate, share data analyses, visualizations and results openly. The design also allows scaling across many instances, the use of could computing and deployment of dashboards and results on production systems. At the same time this environment can be combined with distributed computing to scale not only with the number of users but also with the size of datasets. It has proven to be a fundamental infrastructure for data analytics on large data lakes in the industry. All tools presented in this talk have been released as open source.

Thomas Lumley

Thomas is a biostatistician who has been developing R packages for more than twenty years. He is currently Professor of Biostatistics at the University of Auckland, and teaches in the Statistics and Data Science programs.

DHBins: hexmaps for New Zealand

The DHBins package lets you show statistics for New Zealand Regional Councils and District Health Boards on a hexagonal map, making comparisons easy by size or colour. For categorical data, you can split each hex into six triangles. This was my first ever set of ggplot geoms!

Harel Lustiger

Harel Lustiger is a Data Scientist at Harmonic Analytics Ltd. specialising in Machine Learning Architecture. His work includes building Machine Learning systems for New Zealand’s largest utility companies and public sector organisations.
Harel sees data science as craftsmanship. As such, he is on a continuous journey to find good data science practices. His talks and writings aim to localise and communicate essential practices from Software Engineering and Statistics.
In 2012, after watching IBM Watson competing with (and beating) human players, Harel decided to change his profession from Electrical Engineering to Statistics. Three years later, he became the first Data Science Researcher at the Tel-Aviv University School of Management.

Data Tests: Separating between Data Sources and the Rest

Data science projects in commercial companies often experience a challenge arising from evolving data sources. As the project progresses, new signals and information sources are added incrementally. In practice, when the data source changes, it creates a need to change the application source code.
With no design up front, some application modules, such as a dashboard or a machine learning model, unwittingly become dependent on the data source. In this case, accommodating the evolving data source is not simply a matter of changing the code related to the data source. Rather, preserving the rest of the existing application in a working condition involves further code changes in distant elements of the application.
An alternative way of dealing with evolving data sources is to introduce a small design up front. Such design lets programmers manage the source code dependencies throughout the project life-cycle.
This talk suggests a design that (1) separates data sources from analytic applications and (2) restricts analytic applications from knowing about the data sources.
While the evolving data sources challenge is programming language agnostic, this talk demonstrates an implementation of the suggested design in R.
The takeaway of this talk is a design of a system, where data sources are plugins to the system’s analytic applications.
Presentation Link: https://speakerdeck.com/harell/data-tests

Zhuojia Dai

ZJ has been using R for more than 10 years. He loves contributing to open source.

Handling larger-than-RAM data with ease

{disk.frame} is a new package for manipulating larger-than-RAM data in R. In this talk we will introduce the package and show you how to manipulate 1.8 billion rows of data in under one minute.

Aaron Schiff

Aaron Schiff is a freelance consulting economist and data scientist.

Making your own ggplot2 theme

A brief introduction to creating your own ggplot2 theme, followed by a closer look at some of the fiddly details of making your theme look just so.

Jonathan Ng

Jonathan worked in a data science team in a Fortune 500 company in London where he developed training programs to teach Excel data analysts through to PhD statisticians how to become more productive with the modern R ecosystem. Since then he started his own training and consulting company DataStrategyPro.com which has trained over 12,000 students.

Former Excel User Trains 12,000 Students to Use R

You don’t need to be the most technical person to make a difference. In this talk I’ll discuss my journey, how I went from Excel user to landing a position on a data science team in a Fortune 500 company then starting my own business teaching R to over 12,000 students.

Tom Elliott

Since 2014, Tom has been responsible for most of the development of iNZight, while also working towards a PhD in statistics.

iNZight: behind the scenes

iNZight is an R driven GUI, developed by students at UoA, providing a simple way for statistics learners to visually explore a dataset (and more). In this talk, I’ll give a brief history of the student-led development of iNZight* (and my part in it) before describing a few of the core programming concepts we’ve used (or been stuck with) over the years, including OOP, reactivity, extensibility, and separability; with R examples, of course!
* The GTK-driven desktop version, not the Shiny-driven online version.

Charlotte Jones-Todd

Lecturer, Department of Statistics, University of Auckland. Research focus: spatiotemporal modelling and point processes.

Animating Statistical Theory

Teaching the fundamentals of statistical theory can often be a thankless task. Defining a probability density function or interpreting a hypothesis test? It’s easy to get lost in the formulae. Linking statistical concepts and conveying these links to students can often be challenging. From using a cheeky cat to rejuvenate the linchpin coin toss example to an R shiny app that allows students to explore probability distributions I will discuss making animated plots in R to illustrate some of the fundamentals of statistical theory.

Jayne Ihaka

Jayne is a banker and data enthusiast, with a newfound passion for conversational AI. Jayne works at ANZ New Zealand raising a digital assistant called Jamie.

Chatbot analytics with R

How do you quantify a good conversation? How can analytics help make digital assistants smarter? This talk will cover lessons learned through using R to extract useful insights from conversation data.

Chris Wild

Chris is an Auckland Statistics Department lifer (41 years) and an education, outreach and software enthusiast https://www.stat.auckland.ac.nz/~wild/

iNZight: Capabilities, aims, audiences and ambitions

NZight is an R driven GUI that providing a simple way for visually exploring data and much more. It comes in two “flavours”, a GTK-driven desktop version and a Shiny-driven online version. Our aim is to empower a much wider cross-section of the populace with significant data-science capabilities, making these much more accessible to beginners and occasional users with gentler learning curves and much less reliance on knowing/remembering the names of the things you want to do. We are also working toward closing the GUI-coding loop with system-written R code and markdown files. For example, iNZight already writes code for most of the data-wrangling operations in “R for Data Science”, lots of ggplot2 graphics, and generalised linear modelling. Tom Elliot, the main system architect, is giving a related talk focussing on programming issues.

Emma Vitz

Emma Vitz is an actuary and data scientist with a particular interest in geospatial analysis and data visualisation. At work, Emma spends a lot of her time quantifying the effects of climate change and natural disasters. She loves digging into all kinds of data and communicating what she finds to a wide range of people.

30 Maps in 30 Days: The case for publishing first drafts

In this talk I’ll be chatting about why publishing our first drafts can be so useful. Participating in the 30 Day Map Challenge forced me to produce work quickly and be willing to share first iterations with others. This often felt uncomfortable, but it was also a very useful exercise.

Victoria King

Victoria is a physiology PhD student. She likes making pictures with numbers and good coffee.

I dream of gg(plot2): visualising sleep in pregnancy with R

Sleep is a universal yet highly individual activity. What does it look like in pregnancy? Take a quick look at how simple visualisations can help develop group insights from heterogenous sleep data.

Owen Flanagan

Data Scientist at Mott MacDonald. AUT Masters of Analytics Graduate

What are people tweeting about Donald Trump

An overview of some simple techniques to gather and present a large dataset using AWS and Shiny.

Caleb Moses

Caleb is a Māori mathematician from the Hokianga region whose interests are machine learning, language and automation. Before joining Dragonfly in 2018 he worked at Stats NZ in a statistical analyst role focused on privacy, data visualisation and automating routine tasks. He has a bachelor’s degree and a postgraduate diploma in pure mathematics from the University of Auckland. Caleb is passionate about language. He speaks Japanese fluently and is currently learning Te Reo Māori and Korean.

Reproducible science at Dragonfly

Dragonfly Data Science is a team of 8 scientists who do statistical work across ecology, linguistics, industry and the public sector. We are stubborn proponents of reproducible work, and achieve this using a mix of make, docker and our own continuous integration system, Gorbachev. I will be talking about how we combine these tools to create a flexible, reproducible work environment that we have been able to use to solve a vast range of analytical challenges.

Antony Southworth

Antony started a degree in Mechatronics Engineering at the University of Auckland in early 2014. After stumbling across the data science world, he began studying data science and computer programming in his own time, eventually coming to join Quantiful as Analytics Data Engineer. He now works to close the gap between ‘developer machine’ and ‘production environment’ through automation and containerisation.

R in Production. Yes, really!

A high-level look at how R can be used in production.

John Graves

Data Scientist at Datamine, Senior Data Scientist at Qrious and now Head of Data Science at Electric Kiwi. PhD in Computer Science from AUT. Organiser of Auckland Data Science Lunch http://bit.ly/AKLDSL and partner in Data Science Academy https://www.datascienceacademy.co.nz/

How to be more productive as a Data Scientist

Owen Flanagan recently posted on LinkedIn on this topic. This lightning talk aims to cover his points and more in just 5 minutes.

Afternoon tea

Ending

Q&A and break

Lunch

Morning tea

Registration

Welcome