CODEWITHSUNDEEP
pythonarraysnumpy
Nightingale
Nightingale

Reputation: 143

Is there an Numpy way of combining arrays based on their timestamp row?

By their timestamp, I want to combine 2 arrays with shapes (3, 6) and (3, 8), respectively.

here is the original array:

array([[ 0,  0,  0,  0,  1,  1,  1,  0,  0,  0,  0,  1,  1,  1],
       [ 3,  3,  3,  3,  4,  4,  4,  5,  5,  5,  5,  6,  6,  6],
       [ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14]])

i have extracted the values with ones:

res = np.where(a[0] == 1)
ones = a[:,res].squeeze()
zeros = np.delete(a,res,axis=1)

zeros
(3, 8)
array([[ 0,  0,  0,  0,  0,  0,  0,  0],
       [ 3,  3,  3,  3,  5,  5,  5,  5],
       [ 1,  2,  3,  4,  8,  9, 10, 11]])

ones
(3, 6)
array([[ 1,  1,  1,  1,  1,  1],
       [ 4,  4,  4,  6,  6,  6],
       [ 5,  6,  7, 12, 13, 14]])

Now i want to combine them again by their last row, which is the timestamp.

Is there any numpy solution to that?

Expected result:

array([[ 0,  0,  0,  0,  1,  1,  1,  0,  0,  0,  0,  1,  1,  1],
       [ 3,  3,  3,  3,  4,  4,  4,  5,  5,  5,  5,  6,  6,  6],
       [ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14]])

Upvotes: 0

Views: 77

Answers (2)

Nightingale
Nightingale

Reputation: 143

By using the ideas from both users Thomas and Epsi95, this solved my problem so far. Yet, I'm still open to more performant solutions.

m =np.hstack((zeros, ones))
a = m[:, m[2, :].argsort()]

Upvotes: 2

Thomas Baruchel
Thomas Baruchel

Reputation: 7507

Here is a way (probably not the most compact however):

m = np.hstack((zeros, ones))
m2 = list(zip(m[2], m[0], m[1]))
m2.sort()
np.roll(m2, -1, axis=-1).T

Upvotes: 1

Related Questions