Python: Optimise For loops using list-comprehension

Question

I have only recently started working with Python's list comprehensions. I am not comfortable with them. How would I optimise the code below by using list comprehensions?

heatmap_color = []

for r in xrange(len(heatmap)):
    heatmap_color.append([])
    for c in xrange(len(heatmap[r])):
        heatmap_color[r].append(cmap.to_rgba(heatmap[r][c], alpha=0.5, bytes=True))

What this code does is create a 2D array (representing a colour image's pixel values). The last line maps each one of the existing grayscale pixels in heatmap to a colour with an alpha channel = 0.5 so that these can be entered in the corresponding entry in the heatmap_color array.

Currently it takes just under 30 seconds to run. I would like to reduce this time as much as possible.

Thank you!

Martijn Pieters · Accepted Answer · 2013-07-10 13:35:13Z

4

Nest the loops:

heatmap_color = [[cmap.to_rgba(r, alpha=0.5, bytes=True) for r in c] for c in heatmap]

Note that we don't need to use indices (generated by xrange()); loop directly over the lists instead.

You could use map() and functools.partial() here too to speed things up a little more still:

from functools import partial
heatmap_rgba = partial(cmap.to_rgba, alpha=0.5, bytes=True)
heatmap_color = [map(heatmap_rgba, c) for c in heatmap]

or even:

from functools import partial
heatmap_rgba = partial(cmap.to_rgba, alpha=0.5, bytes=True)
heatmap_per_list = partial(map, heatmap_rgba)
heatmap_color = map(heatmap_per_list, heatmap)

because map() can be faster than list comprehensions, especially when used with functools.partial to avoid the extra stack-push a lambda would require.

edited Jul 10, 2013 at 13:35

answered Jul 10, 2013 at 13:20

Martijn Pieters

1.1m326 gold badges4.2k silver badges3.4k bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Python: Optimise For loops using list-comprehension

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related