Making Data Pretty (and understandable)

A 15-minute presentation to the Python in Astronomy workshop in Seattle. http://python-in-astronomy.github.io/2016/program.html

Transcript

1. Michele Bannister @astrokiwi NRC-Herzberg Astronomy / University of Victoria, BC Making Data Pretty (and understandable)
2. colour palettes what sort of data do you have? chart junk cleaning up plots packages that can make life easier
3. colour palettes sequential diverging qualitative learn more: Cynthia Brewer’s colorbrewer http://colorbrewer2.org/ Try your plots in different perception spaces http://www.vischeck.com/
4. palettable by Matt Davis jiffyclub.github.io/palettable/ pip install palettable > from palettable.colorbrewer.qualitative import Dark2_7
5. http://www.research.ibm.com/people/l/lloydt/color/color.HTM work with the colour response of human eyes
6. work with the colour response of human eyes h"ps://mycarta.wordpress.com/2012/10/06/the-­‐rainbow-­‐is-­‐deadlong-­‐live-­‐the-­‐rainbow-­‐part-­‐3/  
7. matplotlib http://matplotlib.org/users/colormaps.html Defaults are there because something has to be there -flowingdata.com
8. matplotlib http://matplotlib.org/users/colormaps.html jet viridis @astrofrog
9. chart junk “routinely added graphical paraphernalia” - E. Tufte detail impeding understanding
11. https://stanford.edu/~mwaskom/ software/seaborn/tutorial/aesthetics.html
12. heliocentric distance (AU) 0 10 20 30 40 50 inclination (degrees) 0 10 20 30 40 50 inclination (degrees) 1:1 2:1 3:2 5:2 7:3 7:4 5:3 11:4 8:5 15:8 13:5 25 30 35 40 45 50 55 60 65 70 75 80 85 semimajor axis (AU) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 eccentricity 1:1 2:1 3:2 5:2 7:3 7:4 5:3 11:4 8:5 15:8 13:5 classical resonant scattering detached o3e45
13. 0 5 10 15 20 25 30 35 40 inclination (degrees) 38 39 40 41 42 43 44 45 46 47 48 49 semimajor axis (AU) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 eccentricity 2:1 3:2 7:4 centaur classical resonant scattering detached
14. packages for pleasing plots palettable seaborn (has superseded prettyplotlib) for web approaches: d3.js bokeh plotly https://plot.ly/python/#scientific-charts (mpld3 no longer under active development)
15. The greatest value of a picture plot is when it forces us to notice what we never expected to see. J. Tukey, Exploratory Data Analysis, 1977