Problem with plotting ‘scatter3d’ for data with an element number greater than some specific number

I’m exploring ‘Julia v1’ and I run into a problem regarding the ‘scatter3d’ plotting of my data. Interestingly, I don’t get any plot, when I increase the element (column) number over a threshold number 8517. I am wondering causes this?
My code is like this:

using Plots
m = rand(10000, 3)
ThresholderN = 8518
vect = [m[1:ThresholderN, i] for i=1:3]
scatter3d(vect[1], vect[2], vect[3], markersize=1)

Welcome! Please enclose your code in triple backticks (``` like this (multiline possible) ```) to make it more readable.

What package are you using to plot? Do you use a Jupyter notebook, a script, the REPL? Are there any error messages?


Hi! Thank you for the reply.
I made the change regarding the readability. I use Plots, just as I modified in the code. The code gives a plot for numbers below the 8518 and then for numbers equal and above the 8518 it gives nothing, even an error message.

Well, this ones certainly weird - I can reproduce this, and after that I only get send: No error when trying to plot (even basic things).

julia> plot(1:10)
send: No error
send: No error

What’s interesting though is that it kills existing plots (even existing scatter3d). There’s really not much I can debug, but this certainly shouldn’t happen (it already closes everything for me with <=4000 points). I recommend opening an issue at the Plots repo - they’re much more likely to know what’s going on, especially since this seems like the backend is crashing. Please link them this or replicate the information over there.

Is this using the GR backend?

I think so, yes - the behaviour as described happens with the exact code posted by @AlinJulia and if I recall correctly the default backend is GR.

Sorry to resurrect this, but I’m running into the send: No error behavior as well, and this is the only other place I’ve seen that message come up at all. Unfortunately due to a few logistical issues I think it’s going to be difficult for me to create a MWE to reproduce this, but I was hoping that maybe someone here figured something out.