Making better spaghetti (plots): Exploring the individuals in longitudinal data with the brolgar package
Learn more at brolgar.njtierney.com
Slides available here
Video available here
- Longitudinal data as a time series
- Specify structure once
- Use
facet_sample()/facet_strata()to look at data - Summarise with
featuresto find interesting observations - Reconnect summaries to data with a left join
- Repeat this process
Longitudinal (panel) data provide the opportunity to examine temporal patterns of individuals, because measurements are collected on the same person at different, and often irregular, time points. The data is typically visualised using a "spaghetti plot", where a line plot is drawn for each individual. When overlaid in one plot, it can have the appearance of a bowl of spaghetti. With even a small number of subjects, these plots are too overloaded to be read easily. The interesting aspects of individual differences are lost in the noise.
Longitudinal data is often modeled with a hierarchical linear model to capture the overall trends, and variation among individuals, while accounting for various levels of dependence. However, these models can be difficult to fit, and can miss unusual individual patterns. Better visual tools can help to diagnose longitudinal models, and better capture the individual experiences.
In this talk, I introduce the R package, brolgar (BRowse over Longitudinal data Graphically and Analytically in R), which provides tools to identify and summarise interesting individual patterns in longitudinal data.
- Di Cook
- Tania Prvan
- Stuart Lee
- Mitchell O’Hara Wild
- Earo Wang
- Rob Hyndman
- Miles McBain
- Hadley Wickham
- Monash University
- Slides made using xaringan
- Extended with xaringanthemer
- Colours taken + modified from lorikeet theme from ochRe
- Header font is Josefin Sans
- Body text font is Montserrat
- Code font is Fira Mono
Dr. Nicholas Tierney (PhD. Statistics, BPsySci (Honours)) is a Lecturer in Business Analytics and Statistics at Monash University, working with Professors Dianne Cook and Rob Hyndman. His research aims to improve data analysis workflow, and make data analysis more accessible. Crucial to this work is producing high quality software to accompany each research idea. Mostly recently, Nick's work is focussing on exploring longitudinal data (brolgar), and improving how we share data alongside research ( ddd). Other work has focussed on exploring data with the R package visdat, and on creating analysis principles and tools to simplify working with, exploring, and modelling missing data with the package naniar. Nick has experience working with decision trees (treezy), optimisation (maxcovr), Bayesian Data Analysis, and MCMC diagnostics (mmcc.
Nick is a member of the rOpenSci collective, which works to make science open using R, has been the lead organiser for the rOpenSci ozunconf events from 2016-2018 (2016, 2017, 2018), and co-hosts the rstats podcast "Credibly Curious" with Dr. Saskia Freytag. Outside of research, Nick likes to hike, rockclimb, make coffee, bake sourdough, (eventually) knit a hat, take photos, and explore new hobbies.