Data visualisation portfolio
My tenetsâ€‹

Tell the data's story

Keep signaltonoise ratio high

Attention to detail

Codebased for flexibility and reproducibility
â€‹

R statistical environment (base plotting functions)

D3.js (still learning)

Gimp2 and Inkscape (where necessary)
Mapping
Simultaneously show overall model results, interaction effect surfaces, site locations, trends in subgroups and dominant environmental gradient
â€‹
Written in base R with some mapping libraries
100% code
Efficiently show NWSE coastline by rotating map, show site clustering and age chronology models
â€‹
Written in base R with some mapping libraries
100% code
Show seven sites, environmental clustering based on PCA, and plot cluster identity and position along each transect.
â€‹
Written in base R with some mapping libraries
100% code
Show site locations, length of time series and time period.
â€‹
Written in R with rgl package, output as raster.
Show labelled collaborative UQ connections.
â€‹
Written in R with some mapping packages.
99% codebased (some manual movement of numbers in Inkscape).
Data Analysis
Two continuous response variables (x and y axes), each modelled as a function of the same two categorical variables.
â€‹
Written in R, arrow points are custom function I wrote that makes shapes using polygon().
â€‹
98% codebased, some movement of grey ellipses and labels in Inkscape.
Vectors emerging from (0,0), with density and average magnitude summarised by vector angle.
â€‹
Written in R
100% code.
Pair correlation plots of continuous data of interactions with fourlevel categories. Top 10 points on each axis highlighted and labelled. Use of colour gradients to highlight implications of offdiagonal points.
â€‹
Written in R
98% code, labels moved in Inkscape.
Species trait dendrogram with clustering into 8 groups, along with distinguishing trait values for each cluster and mean diversity in four different forest categories.
â€‹
Written in R.
100% code.
Miscellaneous