Lydia R. Lucchesi
On the job market for 2024!
I’m a PhD candidate in the School of Computing at the Australian National University (ANU). My thesis topic is data preprocessing. I’m interested in developing visualisation tools for documenting and communicating data preprocessing decisions. This research is conducted under the supervision of Professor Lexing Xie, Dr. Petra Kuhnert, and Dr. Jenny L. Davis.
I’m a Data61 affiliate at CSIRO (Australia’s national science agency), a team member of the Humanising Machine Intelligence research initiative, and part of the Computational Media Lab at ANU. Prior to ANU, I completed a two-year post-bachelor fellowship at the Institute for Health Metrics and Evaluation (IHME) at the University of Washington, where I modelled the health burden of non-fatal injuries on the Global Burden of Disease study. I have a B.A. in statistics from the University of Missouri and completed my honours thesis on uncertainty visualisation in spatial statistics under the supervision of Professor Christopher K. Wikle.
My research interests include data provenance, data preprocessing, visualisation, uncertainty communication, and reproducible research. I enjoy building visualisation systems in R and being a part of interdisciplinary research teams.
Current projects
The uniting theme between these projects is visualising information, from/about a data analysis, that is often overlooked in research dissemination, despite its relevance to our understanding and interpretation of the analysis.
smallsets: Visual documentation for data preprocessing
The smallsets R package generates visual documentation for data preprocessing in R, R Markdown, Python, and Jupyter Notebooks. Users add structured comments to their preprocessing code and pass it to smallsets. The output from smallsets is a Smallset Timeline, a static and compact visualisation providing a step-by-step explanation of preprocessing decisions. It uses a small set (“Smallset”) of rows from the original dataset to visualise the process at a manageable scale.
smallsets website: lydialucchesi.github.io/smallsets/
Smallset Timeline paper: doi.org/10.1145/3531146.3533175
Below is an example Smallset Timeline visualising the preprocessing of a synthetic dataset with 100 rows and eight columns (C1-C8). It uses a Smallset of five rows and is composed of three Smallset snapshots showing three preprocessing steps.
Vizumap: Visualising uncertainty on maps
The Vizumap R package offers four methods for mapping statistical estimates and uncertainties simultaneously. Users can build a bivariate map, pixel map (with optional animated flickering), glyph map, or exceedance probability map. Vizumap has been used in ecology and public health research. There is also a VizumApp Shiny app based on the Vizumap package.
Below is an example pixel map that I built with Vizumap. Pixel maps divide regions into pixels, and the pixels are filled with values sampled from the confidence interval for a region’s estimate. The U.S. equal area cartogram below visualises 2020 election forecasts from FiveThirtyEight1. The pixel values are uniformly sampled from the 80% confidence interval for each state’s forecasted vote share for Biden. States with wider confidence intervals appear more pixelated because the confidence interval spans more of the colour gradient. The pixel values are sampled from a wider range of colour values, leading to visible colour contrast between the pixels. Conversely, states with narrower confidence intervals appear less pixelated.
1 I used variables
voteshare_chal (vote share), voteshare_chal_lo
(80% CI lower bound), and voteshare_chal_hi (80% CI upper
bound) from the file presidential_state_toplines_2020.csv
available in the FiveThirtyEight
GitHub repository.