Efficient bit-testing conditional iteration over distinct array-elements

Question

I have two 2D-numpy arrays of equal shape, one containing data, and one of type 'ubyte' storing bitflags per pixel. I want to visit every pixel in the data-array that has a particular flag in the bitflags-array.

I could just iterate over every pixel in either array and use the multi-index to get the bitflags and the value of the pixel, e.g.

it = np.nditer(array, flags=['multi_index'])
while not it.finished:
    if bitflags[it.multi_index] & FLAG:
        do_something(array[it.multi_index])
    it.iternext()

Since most pixels do not have the corresponding bitflag set, I would rather like to find all pixels with the given bitflag (for example using numpy.where(bitflags & FLAG) and iterate only over these pixels - something like

pixels = np.where(bitflags & FLAG)
for pixel in array[pixels]:
    do_something()

Is there a way I can still get the indices of the original array to use them in do_something()?

Wolph · Accepted Answer · 2014-10-31 11:55:14Z

1

Not completely sure if I understand your question correctly, but are you looking for this:

pixels, = np.where(bitflags & FLAG)
for i, pixel in zip(pixels, array[pixels]):
    do_something(i, pixel)

edited Oct 31, 2014 at 11:55

answered Oct 31, 2014 at 11:50

Wolph

80.4k12 gold badges142 silver badges152 bronze badges

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

1 Comment

Dux Over a year ago

Such an easy solution! Thank you! Since my arrays are two-dimensional, len(pixels) is 2, so your code does not work directly, but I can get the actual indices using p,y,x in zip(array[pixels], *pixels) I totally missed this, since I rarely need zip

Collectives™ on Stack Overflow

Efficient bit-testing conditional iteration over distinct array-elements

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related