Graphics in Julia

There are multiple packages written for Julia that facilitate graphics. The main difference between their use is the api syntax.

• Winston (Matlab)
• Gaston (gnuplot)
• ASCIIPlots (Exactly what you think it is. I'm not joking.)
• PyPlot (Python matplotlib)
• Gadfly (grammar of graphics)

• Static 2D plotting library for Julia
• Syntax similar to Matlab

using Winston

x = rand(10);
y = rand(10);
plot(x, y, "bo");

savefig("winston_plot.pdf");

In order to print the plot, must save the plot, and insert into a div below.

• Basic plotting functionality
  • scatter plots, line plots, bar charts(ish)
  • plotting functions
  • titles/labels
  • saving (pdf, svg, png, eps)
• Very sparse documentation
• Syntax not very intuitive (to me)
• Limited options/flexibility

• Wrapper for utilizing gnuplot from Julia
• Static 2D plots, 3D surfaces, and image plots

using Gaston

#2D plot
t = 0:0.0001:.15
carrier = cos(2pi*t*200)
modulator = 0.7+0.5*cos(2pi*t*15)
am = carrier.*modulator
plot(t,am,"color","blue","legend","AM DSB-SC","linewidth",1.5,
t,modulator,"color","black","legend","Envelope",
t,-modulator,"color","black","title","AM DSB-SC example",
"xlabel","Time (s)","ylabel","Amplitude",
"box","horizontal top left")
set_filename("plots-julia/gaston_2d.png")
printfigure("png")

#Histogram
y = sqrt( randn(1000).^2 + randn(1000).^2 )
histogram(y,"bins",25,"norm",1,"color","blue","title","Rayleigh density (25 bins)")
set_filename("plots-julia/gaston_histo.png")
printfigure("png")

#3D Surface
gnuplot_send("set view 67,25")
surf(-3:.1:3,-3:0.1:3,(x,y)->cos(x)*sin(y),"plotstyle","points",
"xlabel","coord 1","ylabel","coord 2","zlabel","coord 3",
"title","surf demo","color","blue")
set_filename("plots-julia/gaston_3d.png")
printfigure("png")

• Basic plotting functionality
  • scatter plots, line plots, histograms, surfaces, etc.
  • can overlay histograms/line charts with “mid-level” plotting
  • titles/labels
  • saving (pdf, svg, png, gif)
• Decent documentation (plus documentation for gnuplot is extensive)
• Limited flexibility
• Installation issues (more of a gnuplot issue than Gaston)

using ASCIIPlots
x = rand(10);
y = rand(10);
scatterplot(x, y, sym='*')

  -------------------------------------------------------------
    |           *                                                | 1.05
    |                 **                                         |
    |               *   * *                                      |
    |       ***                                                  |
    |                      *                                     |
    |   *                                                        |
    |                                                            |
    |*                                                           |
    |* *                         *                               |
    |*                            *                             *|
    |                                                            |
    |                                                            |
    |                               *                          * |
    |                                 *                          |
    |                                                            |
    |                                                            |
    |                                     *   *      *           |
    |                                       *    *   *           |
    |                                                            |
    |                                            **              | -1.15
    -------------------------------------------------------------
    0.11                                                    6.17

• Wrapper for utilizing matplotlib (Python) from Julia
• Uses the Julia PyCall package to call matplotlib directly from Julia
• (Sort of/Maybe?) interactive 2D and 3D plots

using PyPlot
x = 0:.5:5
y = exp(-x)

errorbar(x, y, xerr=0.2, yerr=0.4)
savefig("plots-julia/PyPlot_errorbar.svg")

##Scatterplot
N = 50
x = rand(N)
y = rand(N)
colors = rand(N)
area = pi * (15 * rand(N))

scatter(x, y, s=area, c=colors, alpha=0.5)
savefig("plots-julia/PyPlot_scatter.svg")

using PyCall
using PyPlot
@pyimport matplotlib.animation as anim

# First set up the figure, the axis, and the plot element we want to animate
fig = figure()
ax = plt.axes(xlim=(0, 2), ylim=(-2, 2))
global line = ax[:plot]([], [], lw=2)[1]

# initialization function: plot the background of each frame
function init()
    global line
    line[:set_data]([], [])
    return (line,None)
end

# animation function.  This is called sequentially
function animate(i)
    x = linspace(0, 2, 1000)
    y = sin(2 * pi * (x - 0.01 * i))
    global line
    line[:set_data](x, y)
    return (line,None)
end

# call the animator.  blit=True means only re-draw the parts that have changed.
myanim = anim.FuncAnimation(fig, animate, init_func=init,
                               frames=100, interval=20)

myanim[:save]("plots-julia/PyPlots-sinplot.mp4", extra_args=["-vcodec", "libx264", "-pix_fmt", "yuv420p"])

[Source]

• Basic plotting functionality
  • scatter plots, line plots, histograms, surfaces, etc.
  • LaTeX titles/labels
  • saving (png, pdf, ps, eps, svg)
• Very flexible
• Animations!
• Limited documentation (but documentation for matplot is extensive)

• Implementation of the grammar of graphics in Julia
• No longer using D3 in rendering

using Gadfly
using RDatasets
p1 = plot(dataset("datasets", "iris"), x="SepalLength", y="SepalWidth", Geom.point)
p2 = plot(dataset("car", "SLID"), x="Wages", color="Language", Geom.histogram)

draw(SVGJS("plots-julia/gadfly_scatter.js.svg", 6inch, 6inch), p1)
draw(SVGJS("plots-julia/gadfly_hist.js.svg", 6inch, 6inch), p2)

p3 = plot([sin, cos], 0, 25)
draw(SVGJS("plots-julia/gadfly_func.js.svg", 6inch, 6inch), p3)

• Plotting capabilities
  • Layers – Gadfly can draw multiple layers to the same plot
  • LaTeX titles/labels
  • saving (png, pdf, ps, eps, svg , svg-js)
• Very flexible
• Beautiful plots
• Decent documentation
• Slow
• Some interactivity with interact.jl and reactive.jl, but same can be said for all graphics libraries