I need to calculate a mean of values whose indices (row + column) % 5. I understand how to call out specific indices, but I am having trouble defining rows and columns into variables i and j, so that I can create a condition that needs to be met.
data = np.random.normal(1, 0.5, size=(8, 9)) #generate an array
a() # empty list where the values whose indices meet the condition are going to be saved
i, j = np.ix_(data) <- does not work
for x in data
if i + j % 5:
a.add(x)
Your question is not clear about the (row + column) % 5 requirement. But assuming you want the values where (row + column) % 5 == 0 you can build an indexer and make use of advanced indexing.
np.random.seed(1)
data = np.random.normal(1, 0.5, size=(8, 9)) #generate an array
idx = [(i, j)
for i in range(data.shape[0])
for j in range(data.shape[1])
if (i + j) % 5 == 0
]
x = [e[0] for e in idx]
y = [e[1] for e in idx]
filtered = data[(x, y)] # advanced indexing
result = mean(filtered)
print(data)
print(idx)
print(filtered)
print(result)
Output
[[ 1.81217268 0.69412179 0.73591412 0.46351569 1.43270381 -0.15076935
1.87240588 0.61939655 1.15951955]
[ 0.87531481 1.73105397 -0.03007035 0.8387914 0.80797282 1.56688472
0.45005437 0.9137859 0.56107079]
[ 1.02110687 1.29140761 0.44969041 1.57236185 1.45079536 1.25124717
1.45042797 0.65813607 0.93855489]
[ 0.53211528 0.86605596 1.26517773 0.65416962 0.80162324 0.65641365
0.57739718 0.66437693 0.9936677 ]
[ 0.44134483 1.11720785 1.82990109 1.37102208 0.90408222 0.55618552
0.62642085 1.8462273 1.02540388]
[ 0.68150218 1.09545774 2.05012757 1.06007948 1.30860155 1.15008516
0.82387508 0.4287409 0.82532864]
[ 0.89555288 1.2933116 1.41949171 1.46555104 1.14279366 1.44257058
0.62280103 1.62643408 1.25646491]
[ 0.85095358 1.24425907 0.96221414 1.56581469 1.75990841 2.0927877
0.30175183 0.2779431 0.74776707]]
[(0, 0), (0, 5), (1, 4), (2, 3), (2, 8), (3, 2), (3, 7), (4, 1), (4, 6), (5, 0), (5, 5), (6, 4), (7, 3), (7, 8)]
[ 1.81217268 -0.15076935 0.80797282 1.57236185 0.93855489 1.26517773
0.66437693 1.11720785 0.62642085 0.68150218 1.15008516 1.14279366
1.56581469 0.74776707]
0.9958170735553157
Another solution, easier to follow:
import numpy as np
np.random.seed(1)
data = np.random.normal(1, 0.5, size=(8, 9))
a = []
for ix,iy in np.ndindex(data.shape):
if (ix+iy) % 5 == 0:
a.append(data[ix, iy])
mean = np.array(a).mean()
print(a)
print(mean)
ndindex. I was going to propose a similar solution but opted for advanced indexing, that way the new array could be avoided in the end.
(i + j) % 5 == 0?