Matplotlib：显示numpy“稀疏”数组-放大点吗？

Question

I can display numpy array easily like this:

plt.imshow(ary);

the problem is that the array is very sparse and I have to zoom 1-2 times to see "artefacts" or any structure. What will you recommend so that I make this more visible. How would you manipulate the array? May be some trick to enlarge the dots! set nearby pixels in some way?

=================================

here is my first attempt:

lst2 = np.where(ary > 0)
px = zip(lst2[0],lst2[1])
for x,y in px : ary[x-1:x+1, y-1:y+1] = ary[x,y]

---

My second question: would it be possible, If the values in the array are scaled to say range 0-1 or 0-255 or 0-100, to plot them with different colors or alphas in a single command? I can not do:

plt.plot(x,y, '.', markersize=7, color='blue', alpha=ary[x,y] )

because x and y are arrays themselves and alpha/color accept single value only. Is there one line solution or I have to do a loop.

Answer 1

What about

y, x = np.nonzero(ary)
plt.plot(x, y, '.', markersize=5)