unrelated println statement leads to blank plot

I encountered a strange bug in PyPlot that a println statement in a certain place in my code causes a plot to come out completely blank. Below is a stripped-down version of my code. If I `include` the code as is, then it makes three plots, but the third plot window is empty. However, if I comment out the `println` statement on l. 23, then the third plot appears as expected. This is Julia version 1.8.2, PyPlot was just installed today (2022-10-16), python version (according to Pyplot) is 3.9.12, matplotlib version (according to the error message when I type `hold(true)` at the REPL) is 3.5.2, and the platform is Windows 10.
```julia
module driver_ppbug
using LinearAlgebra
using PyPlot

function admm1(a)
n = size(a,2)
clusterID = vcat(ones(Int(n/2)),2*ones(Int(n/2)))
itcount = 1
return clusterID, itcount
end

function find_k_largest_clusters(clusterID, k1)
return vcat([[1,2,3,4], [5,6,7]], [[i] for i = 8 : length(clusterID)])
end

function plotresult(a, clusterID)
n = size(a,2)
colors1 = "rgbcmky"
mxcolor = length(colors1)
result = find_k_largest_clusters(clusterID, mxcolor)
println("result lens = ", [length(q) for q in result])
plotted = [false for i = 1 : n]
for i = 1 : mxcolor
if length(result[i]) > 1
plot(a[1,result[i]], a[2,result[i]], "*" * colors1[[i]])
else
break
end
plotted[result[i]] .= true
end
plot(a[1, .!plotted], a[2, .!plotted], "." * colors1[[mxcolor]])
axis("equal")
nothing
end

function gm_generator(n)
a = hcat(collect(1.0:1.0:n), sin.(1.0:1.0:n))'
compID = vcat(ones(Int(n/2)), 2*ones(Int(n/2)))
weight = ones(n)
return a, compID, weight
end

function mxgau(n, sigma, lambdascale1, lambdascale2)
a, compID, weight = gm_generator(n)
sub1 = findall(compID .== 1)
sub2 = findall(compID .== 2)
close("all")
figure(1)
plot(a[1,sub1], a[2,sub1], "*r")
plot(a[1,sub2], a[2,sub2], "*g")
axis("equal")
figure(2)
plot(a[1,:], a[2,:], "*k")
axis("equal")
clusterID, itcount1 = admm1(a)
figure(3)
plotresult(a, clusterID)
println("itcount1 = ", itcount1)
nothing
end

mxgau(20, .3, .3, .1)

end

```

Reproduce the reported behavior by running the module and calling mxgau(20, .3, .3, .1) on the stated Julia, PyPlot, Python, and matplotlib versions. Start with plotresult and the figure calls in mxgau, then compare runs with and without the println; done means explaining the blank third figure and verifying that all three plots render correctly.

由索引模型根据 Issue 内容生成。