Iām not sure what you mean with āfrom the REPLā but it seems that Makie would be good for this since it should update the canvas quickly and handle mouse events.
What I mean is that I donāt want to run Julia from Juno or Atom or Jupyter, I want to interactively plot from the REPL in the standard unix environment without using any web-browser interfaces.
Can Makie be used without the web browser interface? Iād prefer something like a Qt popup window.
There are two Mandelbrot examples and one Julia set example in the examples section of GR.jl - two plain Julia scripts (mandel.jl, julia.jl) and an OpenCL accelerated Mandelbrot script.
I already know I can plot fractals in plotting packages.
To answer my question you would also have to take the interactivity into account, like mouse hover and click and drag and so onā¦ can they do that? Please show how I would make an interactive plot that updates with click and drag coordinate parametric interactivity.
For example, placing dots into the plot and clicking and dragging them into to new configurations. The plot should be re-computed with the new configurations.
Yeah, Gaston (or rather, gnuplot) is not designed to do that. It wasnāt clear to me from your question and comments what you meant by āinteractiveā.
InspectDR written in Julia, so you can add interactivity if you wish.
It was not designed to hook in custom callbacks yet, but you can use ādynamic patchingā (guerrilla patch/monkey patch) to āhackā in your own hooks at the moment.
You can find an example of this here:
Note that it might be more appropriate to overwrite the callback functions instead of the handleevent_mousepress() function. I can guide you through that if you want.
Limitations
InspectDR does not support heat map or contour plots - so you probably donāt want to use it for drawing fractals. But it does work fine for 2D line plots.
Layered design
InspectDR typically is used as a plot application, but you could also instantiate an individual InspectDR plot widget within your custom, interactive Gtk application.
ā¦But if you want to do that, you might want an overview of how the code was divided into files. I will only do that if you ask, I donāt feel like explaining this for nothing.