I’m trying to reproduce this bar plot:
With the following code:
using Plots,DelimitedFiles
f = readdlm("ch3_data_english.txt")
bar(f/100,
xticks = (1:26,'a':'z'),
yticks = 0:0.02:0.2,
ylims = (0,0.13),xlims = (1,26),
color = RGB(0.9,0.6,0),
label = false)
Which output is
My only complaint is that the little lines above the letters don’t show up, how do I fix that?
That would be an interesting feature. I discovered that the ticks show up reducing the alpha argument, there is a way to plot the bars behind the ticks ?
Changing the backend to PyPlot and a:z to string.(a:z) the result is:
That’s almost perfect, now I only need the ticks to appear.
The Box you can get with GR with framestyle=:box.
In case you don’t want to change the default backend (I noticed that this is one example of that book).
What’s the use of those ticks? There is one bar per label, and they are in direct connection. The fact that Plots.jl removes them is good-actually.
I don’t think they r dropped, just they r being at the background. at q there is a small tick
Use the plots argument: tick_direction=:out, to get the ticks outside the box.
It is probably not needed to reproduce exactly the same plot layout in reproducing those examples of the book. I guess the python and matlab codes that are shown do not produce identical plots either (keeping the code simple is likely a preferable goal).
Speaking of xticks, I am wondering if the docs of Plots.jl concerning the attributes are not wrong. Here, I can see xsticks instead of xticks. Is it a mistake?
That seems to be a line type (“sticks along x”) for gaston() and pgfplotsx() backends. Not the same thing as xticks attribute:
using Plots; pgfplotsx()
plot(rand(10), lt=:xsticks)
That’s true, the pyhon code certainly do not reproduce a identical plot, I will choose a more simple approach.
