I was playing around with using Gadfly's Geom.ribbon to make plots showing
integrals as shaded areas under curves. Sometimes you might have two curves
in separate colors where parts of the shaded regions under both curves
overlap. If the shading that is being done by Geom.ribbon is not completely
opaque, you should then get an overlap region where the shading color is a
blend of the two original ones. I thought this behavior would result by
default as a similar thing is what appears to be going on in the example
for Geom.ribbon in the (old) Gadfly doc here:
https://dcjones.github.io/Gadfly.jl/geom_ribbon.html. But the following
example I tried gives this instead:
using Gadfly, DataFrames, Distributions
d1 = Normal(-1); d2 = Normal(1);
x=-4:0.01:4
y1 = pdf(d1,x); y2 = pdf(d2,x);
df1 = DataFrame(x=x,y=y1,ymin=0.0,ymax=y1,d="d1");
df2 = DataFrame(x=x,y=y2,ymin=0.0,ymax=y2,d="d2");
df = vcat(df1,df2)
p = plot(df, x=:x, y=:y, ymin=:ymin, ymax=:ymax, color=:d, Geom.line, Geom.
ribbon)
draw(PNG("test1.png", 12cm, 6cm), p)
<https://lh5.googleusercontent.com/-v018Iz5Xie4/VO6az8VLc1I/AAAAAAAAAAM/w8ZPou8r8KU/s1600/test1.png>
Changing what I believe is the relevant Theme parameter to explicitly
specify an opacity doesn't change the output:
p = plot(df, x=:x, y=:y, ymin=:ymin, ymax=:ymax, color=:d, Geom.line, Geom.
ribbon, Theme(lowlight_opacity=0.5))
I then noticed that the same Geom.ribbon example on the new Gadfly site
also doesn't have the transparency effect the old example did:
http://gadflyjl.org/geom_ribbon.html.
So opacity level settings are documented and appear to have been working at
some point but aren't now. Anyone happen to know what the situation with
this is?
