Plots.jl click input

Click somewhere on this graph:

xxxxxxxxxx

9.5 μs

xxxxxxxxxx
 
@initially [5,5] @bind x0 plotclicktracker(linepoint(x0))

11.2 s

Int641

xxxxxxxxxx
 
x0

200 ns

linepoint (generic function with 1 method)

xxxxxxxxxx
 
function linepoint(click_coordinate)
    p = plot(1:10)
    scatter!(p, click_coordinate[1:1], click_coordinate[2:2])
end

22.1 μs

Why is this special? Because normally, the input (@bind) and output live in two separate cells:

xxxxxxxxxx

7.7 μs

xxxxxxxxxx
 
@bind x html"<input type=range>"

12.2 ms

missing

xxxxxxxxxx
 
sqrt(x)

100 ns

But in our interactive plot, the input and output are inside the same cell. To achieve this, we need a couple of tricks: a click tracker wrapper, a self-referencing cell and the @initially macro. Let's talk about those one by one:

Click tracker

TUTORIAL WORK IN PROGRESS

xxxxxxxxxx

13.9 μs

COOL_PLOT (generic function with 1 method)

xxxxxxxxxx
 
function COOL_PLOT(q0)
    p = plot(1:10)
    scatter!(p, q0[1:1], q0[2:2])
    
    return p
end

23.3 μs

The third step is to write a function that:

takes a click coordinate (a vector [x, y]) as input, and
returns a plot.

It's up to you to use this coordinate in a creative way!

xxxxxxxxxx

15.9 μs

Necessary functions

You can copy these cells into your notebook by selecting them (start dragging a selection rectangle from inbetween two cells), and dragging them into another notebook.

xxxxxxxxxx

5.1 μs

plotclicktracker (generic function with 1 method)

xxxxxxxxxx

92.6 μs

intially_function (generic function with 1 method)

xxxxxxxxxx

65.7 μs

You can also use ingredients to load this notebook into another notebook.

xxxxxxxxxx

6.5 μs

Packages

(We only need Plots.jl)

xxxxxxxxxx

4.4 μs

xxxxxxxxxx
 
begin
    import Pkg
    Pkg.activate(mktempdir())
    Pkg.add([
            Pkg.PackageSpec(name="Plots", version="1.6-1"),
            ])
    using Plots
end

7.1 s