matplotlib plt.show() only selected objects

Question

I have several plt.plot instances and I wanted to only plt.show() certain objects. To illustrate here is some code:

import matplotlib.pyplot as plt

ax1 = plt.plot(range(5),range(5))
ax2 = plt.plot([x+1 for x in range(5)],[x+1 for x in range(5)])
ax3 = plt.plot([x+2 for x in range(5)],[x+2 for x in range(5)])

#plt.show([ax1,ax2])
plt.show()

So I would like something like the commented out statement, to display only ax1 & ax2 in the example figure.

Answer 1

You can remove some of the plotted lines from the set of lines of the current axes:

axes = plt.gca()  # Get current axes
axes.lines.remove(ax2[0])  # Removes the (first and only) line created in ax2
plt.draw()  # Updates the graph (in interactive mode)

If you want to put it back, you can similarly do

axes.lines.append(ax2[0])  # Puts the line back (the drawing order is changed, here)

You could alternatively save the current graph lines, if you need to put them back later:

all_lines = list(axes.lines)  # Copy
# ...
axes.lines[:] = all_lines  # All lines put back

The key point is that each plot() command adds a line to the current axes and draws it (in interactive mode). So you can either remove already plotted lines (like in this answer).

As Yann pointed out, you can also make some lines invisible. However, the method from this answer is probably faster, since there are fewer lines to be drawn (if this matters).

Answer 2

Not exactly. First off, the plt.plot call does not return an axes, it returns a list of Line2D objects, one for each line plotted. You can use the OO interface to Matplotlib to create a separate axes for each plot, then selectively add those as subplots, etc. There are a lot of different ways to selectively reveal a plot.

But for your example, you can take advantage of the Line2D's alpha value, ie how opaque it is, to make any one line invisible. Here's a modified version of your example:

import matplotlib.pyplot as plt

line1 = plt.plot(range(5),range(5))
line2 = plt.plot([x+1 for x in range(5)],[x+1 for x in range(5)])
line3 = plt.plot([x+2 for x in range(5)],[x+2 for x in range(5)])
print line3, " see I'm a list of lines"
print line3[0].get_alpha()
line3[0].set_alpha(0) # make complete opaque

#plt.show([ax1,ax2])
plt.gcf().savefig('line3opaque.png')
line3[0].set_alpha(1) # make visible
line1[0].set_alpha(0) # make opaque
plt.gcf().savefig('line1opaque.png')
plt.show()

The first figure I saved is 'line3opaque.png'; this is what I get:

Line 3 is not there and lines 1 and 2 are. For 'line1opaque.png' I get:

Now we have line 1 missing.