LET YOUR DATA SPEAK! DATA VISUALISATION IN MATPLOTLIB by Bartosz Teleńczuk
Jul 15, 2015
LET YOUR DATA SPEAK!DATA VISUALISATION IN MATPLOTLIB
by Bartosz Teleńczuk
INTR
ODUC
TION
ANSCOMBE DATASET
EXPLORATORY DATA ANALYSISThe greatest value of a picture is when it forces
us to notice what we never expected to see.
VISUAL PRESENTATION
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
The Journal of neuroscience :...Nature neurosciencePloS oneNatureNeuronProceedings of the National...Nature methodsCurrent opinion in...Science (New York, N.Y.)Current biology : CBJournal of neurophysiologyFrontiers in neural circuitsBrain researchFrontiers in molecular...The Journal of biological...Conference proceedings : ......Progress in brain researchJournal of neuroscience...Journal of visualized...Neuroscience researchCellBiological psychiatryThe Journal of physiologyNature protocolsBehavioural brain researchExperimental physiologyNeuropharmacologyNeuroscienceNature communicationsNeurosurgery
INTERACTIVE
PRESENTATION
John Hunter (1968 - 2012)
PATTERNS OVER TIME
BACKENDSTkAgg Qt4Agg WebAgg
ANTIGRAIN GEOMETRY
Raster (PNG) Vector (SVG)
PYPLOT INTERFACE import matplotlib.pyplot as pltplt.figure()plt.subplot(...)plt.plot(...) plt.show()
pyplot is a stateful interface that handles muchof the boilerplate for creating figures and axesand connecting them to the backend of your
choice, and maintains module-level internal datastructures representing the current figure and
axes to which to direct plotting commands.
RECAPplt.plotplt.showplt.xlabelplt.ylabelplt.xlimplt.gcaplt.text
xaxis.set_minor_locatoryaxis.set_major_locator
YearsLocatorMonthLocator
PROPORTIONS
ARTISTS
Figure Axes
Line2D
X/YAxis
Spines
Label
Ticks
DESIGN POINTS
SIGNAL-TO-NOISE RATIO
RECAPplt.bar
plt.xticks(pos, labels) Spine.set_visible
Axis.set_ticks_position Axes.set_axis_bgcolor
plt.grid plt.title
DISTRIBUTIONS
STEM-AND-LEAF PLOT1.1 2.2 2.8 3.2 3.4 3.4 3.5 4.1 4.9 5.5 5.7
1. 12. 2 83. 2 4 4 54. 1 95. 5 7
BOXPLOTS
TRANSFORMS
data transform
axes transform
(7,1)
(7,-1)(0,-1)
(0,1)
(1,0)
(1,1)(0,1)
(0,0)
figure transform (1,1)(0,1)
(0,0) (1,0)
RECAPplt.histplt.boxplotplt.legendplt.textplt.setpplt.vlines
Axes.transAxesAxes.transData
blended_transform_factory
CORRELATIONS
SMALL MULTIPLIES
USE OF COLOR
*
*
jet
hot
gray
heatmap
Bang Wong, Nature Methods, 2010
VISUAL ILLUSION
MATPLOTLIB COLORMAPS
RECAPplt.subplots
Axes.set_axis_offplt.imshowrcParams
FINDING PATTERNS
PRINCIPAL COMPONENT ANALYSIS
K-MEANS CLUSTERING
VISUAL ENCODING CHANNELSposition
color saturation
length
angle
area
volume
color intensity
shape
color hue
RECAPplt.scatter
mcolors.Normalizecm.get_cmap
MAKING MAPS
MAP PROJECTIONS
CHOROPLETH MAP
RECAPmatplotlib plt.contourf cartopy
cartopy.crscartopy.ioax.coastlinesax.add_featureax.set_extentax.add_image
ax.add_geometries