Visualisation
Table of Contents
Notes
Miter joins
This is what you want to search when you are annoyed by line segments not meeting up cleanly when you have thicker lines.
File formats
| Abbreviation | Full name | Vector/Raster |
|---|---|---|
| EPS | Encapsulated PostScript | Vector |
| Portable Document Format | Vector | |
| SVG | Scalable Vector Graphics | Vector |
| TIFF | Tag Image File Format | Raster |
inkscape and convert are two command line tools for converting between
vector and raster formats respectively.
Paper sizes
| Paper | Width | Height |
|---|---|---|
| A4 | 210mm | 297mm |
| A5 | 148mm | 210mm |
| A6 | 105mm | 148mm |
| A7 | 74mm | 105mm |
Font sizes
For figures, the axis and legend titles should be 10pt (not italic or bold), the axis and legend text should be 9pt. Setting this in pt will make the results consistent provided you save to the true size of the figure.
You can comfortably fit a 1x3 grid of plots on A4 with text of size 9–10pt, if each plot is of size 50x60mm.
Colour
ColorBrewer2 palettes
You can download a GIMP colour palette .gpl files and load them into
your editor of choice:
Colour picker
Colour blindness
For R users who want a drop-in solution for avoiding issues with limited colour perception there is colorBlindness, a package which automates the palette selection process.
Simple R colours
Here is a little snippet of R code that makes it easier to choose some basic colours.
## Define some hex codes to use as a colour scheme. These ## are from the ColorBrewer2 project: ## \code{https://colorbrewer2.org/} my_colours <- list( truth = list(dark = "#d73027", light = "#fee090"), estimate = list(dark = "#4575b4", light = "#e0f3f8"), other = list( grey = list( light = "#e0e0e0", medium = "#999999", dark = "#4d4d4d" ) ) )
Curated palettes
- The ultimate resource for visualisations is colorbrewer2.
Solarized
WEBSITE of Ethan Schoonover the creator
| Colour | RGB |
|---|---|
| base03 | #002b36 |
| base02 | #073642 |
| base01 | #586e75 |
| base00 | #657b83 |
| base0 | #839496 |
| base1 | #93a1a1 |
| base2 | #eee8d5 |
| base3 | #fdf6e3 |
| yellow | #b58900 |
| orange | #cb4b16 |
| red | #dc322f |
| magenta | #d33682 |
| violet | #6c71c4 |
| blue | #268bd2 |
| cyan | #2aa198 |
| green | #859900 |
Here is a snippet to add to your CSS if you want to use it.
:root { --base03: #002b36; --base02: #073642; --base01: #586e75; --base00: #657b83; --base0: #839496; --base1: #93a1a1; --base2: #eee8d5; --base3: #fdf6e3; --yellow: #b58900; --orange: #cb4b16; --red: #dc322f; --magenta: #d33682; --violet: #6c71c4; --blue: #268bd2; --cyan: #2aa198; --green: #859900; }
Or if you want to use it in LaTeX
\usepackage{xcolor} \definecolor{base03}{HTML}{002b36} \definecolor{base02}{HTML}{073642} \definecolor{base01}{HTML}{586e75} \definecolor{base00}{HTML}{657b83} \definecolor{base0}{HTML}{839496} \definecolor{base1}{HTML}{93a1a1} \definecolor{base2}{HTML}{eee8d5} \definecolor{base3}{HTML}{fdf6e3} \definecolor{yellow}{HTML}{b58900} \definecolor{orange}{HTML}{cb4b16} \definecolor{red}{HTML}{dc322f} \definecolor{magenta}{HTML}{d33682} \definecolor{violet}{HTML}{6c71c4} \definecolor{blue}{HTML}{268bd2} \definecolor{cyan}{HTML}{2aa198} \definecolor{green}{HTML}{859900}
Links
Materials
- SVG Repo is a collection of free icons
- Healthicons is a collection of free health related images
- Bioicons is a collection of free biology related images
Tools
- colour converter page for moving between representations.
- Vega visualisation grammar
- Vega-Lite
- Voyager a tool for rapid EDA
- Vega Cookbook
- Dithering tools
- colorbrewer2
- colormaps for matplotlib
- ggplot2 extensions
- Medoid as a way to select the middle of a dataset
- Nomograms can be generated using PyNomo.
- WebPlotDigitizer for extracting data from figures
References
- Scientific Visualization: Python + Matplotlib is an open source book
- Effective Figure Design course
- Babraham Bioinformatics course: Scientific Figure Design which includes some very nice notes on using Inkscape
- DataVisProject: a catalogue of visualisation types
- R Graph Gallery
- R spatial blog posts
- University of Melbourne branding guide V1
- Guide to designing figures from Nature Reviews