I'm using np.nonzero() and i dont understand the return
I try
for groupPosition in np.nonzero(groupMatrix):
print groupPosition
and return [0 0 1 2 3 3 3]
for groupPosition in zip(np.nonzero(groupMatrix)):
print groupPosition
and return (array([0, 1, 0, 3, 0, 1, 3]),)
groupMatrix:
[[ 1. 1. 0. 0.]
[ 1. 0. 0. 0.]
[ 0. 0. 0. 2.]
[ 3. 3. 0. 2.]]
But don't return the position like a (0, 0)
>>> import numpy as np
>>>
>>> groupMatrix = np.array([
... [1, 1, 0, 0],
... [1, 0, 0, 0],
... [0, 0, 0, 2],
... [3, 3, 0, 2]
... ])
>>> np.nonzero(groupMatrix)
(array([0, 0, 1, 2, 3, 3, 3], dtype=int64), array([0, 1, 0, 3, 0, 1, 3], dtype=int64))
>>> zip(np.nonzero(groupMatrix))
[(array([0, 0, 1, 2, 3, 3, 3], dtype=int64),), (array([0, 1, 0, 3, 0, 1, 3], dtype=int64),)]
Use zip(*...):
>>> zip(*np.nonzero(groupMatrix))
[(0, 0), (0, 1), (1, 0), (2, 3), (3, 0), (3, 1), (3, 3)]
zip(*a) is like zip(a[0], a[1], ...)
>>> a = [(0, 1, 2), (3, 4, 5)]
>>> zip(a)
[((0, 1, 2),), ((3, 4, 5),)]
>>> zip(a[0], a[1])
[(0, 3), (1, 4), (2, 5)]
>>> zip(*a)
[(0, 3), (1, 4), (2, 5)]
Try the following:
import numpy as np
var = [
[1.0, 1.0, 0.0, 0.0],
[1.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 2.0],
[3.0, 3.0, 0.0, 2.0]
]
rows, cols = np.nonzero(var)
for r, c in zip(rows, cols):
print var[r][c]
Returns:
1.0
1.0
1.0
2.0
3.0
3.0
2.0
You are getting the results you are getting because np.nonzero returns a tuple, since your array has 2 dimensions, it has two arrays. Now, each of these arrays need to be used together, so in my example, the function returns the row number, and then the column number. Lets have a look see:
>>> import numpy as np
>>> var = [
[1.0, 1.0, 0.0, 0.0],
[1.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 2.0],
[3.0, 3.0, 0.0, 2.0]
]
>>>
>>> non_zeroes = np.nonzero(var)
>>> non_zeroes
(array([0, 0, 1, 2, 3, 3, 3]), array([0, 1, 0, 3, 0, 1, 3]))
If we take a close look, we can see that var[0][0] is indeed non zero. so is var[3][3]. However, you will not see a 2 in the first tuple and another 2 at the corresponding index.