Invert binary numpy array

Question

is there a kind of "logical no" for numpy arrays (of numbers of course).

For example, consider this array: x = [1,0,1,0,0,1]

i am looking for an easy way to compute its "inverse" y = [0,1,0,1,1,0]

yatu · Accepted Answer · 2019-02-28 08:57:52Z

For an array of 1s and 0s you can simply subtract the values in x from 1:

x = np.array([1,0,1,0,0,1]) 1-x # array([0, 1, 0, 1, 1, 0])

Or you could also take the bitwise XOR of the binary values in x with 1:

x^1 # array([0, 1, 0, 1, 1, 0])

Scott Boston · Accepted Answer · 2019-02-27 14:26:40Z

5

Yes, you can use np.logical_not:

np.logical_not(x).astype(int)

Output:

array([0, 1, 0, 1, 1, 0])

answered Feb 27, 2019 at 14:26

154k15 gold badges160 silver badges207 bronze badges

Comments

Psytho · Accepted Answer · 2019-02-27 14:34:01Z

5

Or using XOR:

y = [n ^ 1 for n in x]

answered Feb 27, 2019 at 14:34

3,4142 gold badges22 silver badges28 bronze badges

fountainhead · Accepted Answer · 2019-02-27 15:07:42Z

Here's one way:

y = (x == 0).astype(int)

Alternatively:

y = 0 + (x == 0)

Output:

[0 1 0 1 1 0]

Notes:

(x == 0) gives a boolean array where False appears in the place of 1, and True appears in the place of 0.
Calling the method astype(int), or adding scalar 0 to the matrix, converts False to 0 and True to 1