Related
I have two arrays A and i with dimensions (1, 3, 3) and (1, 2, 2) respectively. I want to define a new array I which gives the elements of A based on i. The current and desired outputs are attached.
import numpy as np
i=np.array([[[0,0],[1,2],[2,2]]])
A = np.array([[[1,2,3],[4,5,6],[7,8,9]]], dtype=float)
I=A[0,i]
print([I])
The current output is
[array([[[[1.000000000, 2.000000000, 3.000000000],
[1.000000000, 2.000000000, 3.000000000]],
[[4.000000000, 5.000000000, 6.000000000],
[7.000000000, 8.000000000, 9.000000000]],
[[7.000000000, 8.000000000, 9.000000000],
[7.000000000, 8.000000000, 9.000000000]]]])]
The desired output is
[array(([[[1],[6],[9]]]))
In [131]: A.shape, i.shape
Out[131]: ((1, 3, 3), (1, 3, 2))
That leading size 1 dimension just adds a [] layer, and complicates indexing (a bit):
In [132]: A[0]
Out[132]:
array([[1, 2, 3],
[4, 5, 6],
[7, 8, 9]])
This is the indexing that I think you want:
In [133]: A[0,i[0,:,0],i[0,:,1]]
Out[133]: array([1, 6, 9])
If you really need a trailing size 1 dimension, add it after:
In [134]: A[0,i[0,:,0],i[0,:,1]][:,None]
Out[134]:
array([[1],
[6],
[9]])
From the desired numbers, I deduced that you wanted to use the 2 columns of i as indices to two different dimensions of A:
In [135]: i[0]
Out[135]:
array([[0, 0],
[1, 2],
[2, 2]])
Another way to do the same thing:
In [139]: tuple(i.T)
Out[139]:
(array([[0],
[1],
[2]]),
array([[0],
[2],
[2]]))
In [140]: A[0][tuple(i.T)]
Out[140]:
array([[1],
[6],
[9]])
You must enter
I=A[0,:1,i[:,1]]
You can use numpy's take for that.
However, take works with a flat index, so you will need to use [0, 5, 8] for your indexes instead.
Here is an example:
>>> I = [A.shape[2] * x + y for x,y in i[0]] # Convert to flat indexes
>>> I = np.expand_dims(I, axis=(1,2))
>>> A.take(I)
array([[[1.]],
[[6.]],
[[9.]]])
The output of the two commands below gives a different array shape, I do appreciate explaining why and referring me to a reference if any, I searched the internet but did not find any clear explanation for it.
data.shape
(11,2)
# outputs the values in column-0 in an (1x11) array.
data[:,0]
array([-7.24070e-01, -2.40724e+00, 2.64837e+00, 3.60920e-01,
6.73120e-01, -4.54600e-01, 2.20168e+00, 1.15605e+00,
5.06940e-01, -8.59520e-01, -5.99700e-01])
# outputs the values in column-0 in an (11x1) array
data[:,:-1]
array([[-7.24070e-01],
[-2.40724e+00],
[ 2.64837e+00],
[ 3.60920e-01],
[ 6.73120e-01],
[-4.54600e-01],
[ 2.20168e+00],
[ 1.15605e+00],
[ 5.06940e-01],
[-8.59520e-01],
[-5.99700e-01]])
I'll try to consolidate the comments into an answer.
First look at Python list indexing
In [92]: alist = [1,2,3]
selecting an item:
In [93]: alist[0]
Out[93]: 1
making a copy of the whole list:
In [94]: alist[:]
Out[94]: [1, 2, 3]
or a slice of length 2, or 1 or 0:
In [95]: alist[:2]
Out[95]: [1, 2]
In [96]: alist[:1]
Out[96]: [1]
In [97]: alist[:0]
Out[97]: []
Arrays follow the same basic rules
In [98]: x = np.arange(12).reshape(3,4)
In [99]: x
Out[99]:
array([[ 0, 1, 2, 3],
[ 4, 5, 6, 7],
[ 8, 9, 10, 11]])
Select a row:
In [100]: x[0]
Out[100]: array([0, 1, 2, 3])
or a column:
In [101]: x[:,0]
Out[101]: array([0, 4, 8])
x[0,1] selects an single element.
https://numpy.org/doc/stable/user/basics.indexing.html#single-element-indexing
Indexing with a slice returns multiple rows:
In [103]: x[0:2]
Out[103]:
array([[0, 1, 2, 3],
[4, 5, 6, 7]])
In [104]: x[0:1] # it retains the dimensions, even if only 1 (or even 0)
Out[104]: array([[0, 1, 2, 3]])
Likewise for columns:
In [106]: x[:,0:1]
Out[106]:
array([[0],
[4],
[8]])
subslices on both dimensions:
In [107]: x[0:2,1:3]
Out[107]:
array([[1, 2],
[5, 6]])
https://numpy.org/doc/stable/user/basics.indexing.html
x[[0]] also returns a 2d array, but that gets into "advanced" indexing (which doesn't have a list equivalent).
I've seen variations of this question asked a few times but so far haven't seen any answers that get to the heart of this general case. I have an n-dimensional array of shape [a, b, c, ...] . For some dimension x, I want to look at each sub-array and find the coordinates of the maximum.
For example, say b = 2, and that's the dimension I'm interested in. I want the coordinates of the maximum of [:, 0, :, ...] and [:, 1, :, ...] in the form a_max = [a_max_b0, a_max_b1], c_max = [c_max_b0, c_max_b1], etc.
I've tried to do this by reshaping my input matrix to a 2d array [b, a*c*d*...], using argmax along axis 0, and unraveling the indices, but the output coordinates don't wind up giving the maxima in my dataset. In this case, n = 3 and I'm interested in axis 1.
shape = gains_3d.shape
idx = gains_3d.reshape(shape[1], -1)
idx = idx.argmax(axis = 1)
a1, a2 = np.unravel_index(idx, [shape[0], shape[2]])
Obviously I could use a loop, but that's not very pythonic.
For a concrete example, I randomly generated a 4x2x3 array. I'm interested in axis 1, so the output should be two arrays of length 2.
testarray = np.array([[[0.17028444, 0.38504759, 0.64852725],
[0.8344524 , 0.54964746, 0.86628204]],
[[0.77089997, 0.25876277, 0.45092835],
[0.6119848 , 0.10096425, 0.627054 ]],
[[0.8466859 , 0.82011746, 0.51123959],
[0.26681694, 0.12952723, 0.94956865]],
[[0.28123628, 0.30465068, 0.29498136],
[0.6624998 , 0.42748154, 0.83362323]]])
testarray[:,0,:] is
array([[0.17028444, 0.38504759, 0.64852725],
[0.77089997, 0.25876277, 0.45092835],
[0.8466859 , 0.82011746, 0.51123959],
[0.28123628, 0.30465068, 0.29498136]])
, so the first element of the first output array will be 2, and the first element of the other will be 0, pointing to 0.8466859. The second elements of the two matrices will be 2 and 2, pointing to 0.94956865 of testarray[:,1,:]
Let's first try to get a clear idea of what you are trying to do:
Sample 3d array:
In [136]: arr = np.random.randint(0,10,(2,3,4))
In [137]: arr
Out[137]:
array([[[1, 7, 6, 2],
[1, 5, 7, 1],
[2, 2, 5, *6*]],
[[*9*, 1, 2, 9],
[2, *9*, 3, 9],
[0, 2, 0, 6]]])
After fiddling around a bit I came up with this iteration, showing the coordinates for each middle dimension, and the max value
In [151]: [(i,np.unravel_index(np.argmax(arr[:,i,:]),(2,4)),np.max(arr[:,i,:])) for i in range
...: (3)]
Out[151]: [(0, (1, 0), 9), (1, (1, 1), 9), (2, (0, 3), 6)]
I can move the unravel outside the iteration:
In [153]: np.unravel_index([np.argmax(arr[:,i,:]) for i in range(3)],(2,4))
Out[153]: (array([1, 1, 0]), array([0, 1, 3]))
Your reshape approach does avoid this loop:
In [154]: arr1 = arr.transpose(1,0,2) # move our axis first
In [155]: arr1 = arr1.reshape(3,-1)
In [156]: arr1
Out[156]:
array([[1, 7, 6, 2, 9, 1, 2, 9],
[1, 5, 7, 1, 2, 9, 3, 9],
[2, 2, 5, 6, 0, 2, 0, 6]])
In [158]: np.argmax(arr1,axis=1)
Out[158]: array([4, 5, 3])
In [159]: np.unravel_index(_,(2,4))
Out[159]: (array([1, 1, 0]), array([0, 1, 3]))
max and argmax take only one axis value, where as you want the equivalent of taking the max along all but one axis. Some ufunc takes a axis tuple, but these do not. The transpose and reshape may be the only way.
In [163]: np.max(arr1,axis=1)
Out[163]: array([9, 9, 6])
I'd like to copy a numpy 2D array into a third dimension. For example, given the 2D numpy array:
import numpy as np
arr = np.array([[1, 2], [1, 2]])
# arr.shape = (2, 2)
convert it into a 3D matrix with N such copies in a new dimension. Acting on arr with N=3, the output should be:
new_arr = np.array([[[1, 2], [1,2]],
[[1, 2], [1, 2]],
[[1, 2], [1, 2]]])
# new_arr.shape = (3, 2, 2)
Probably the cleanest way is to use np.repeat:
a = np.array([[1, 2], [1, 2]])
print(a.shape)
# (2, 2)
# indexing with np.newaxis inserts a new 3rd dimension, which we then repeat the
# array along, (you can achieve the same effect by indexing with None, see below)
b = np.repeat(a[:, :, np.newaxis], 3, axis=2)
print(b.shape)
# (2, 2, 3)
print(b[:, :, 0])
# [[1 2]
# [1 2]]
print(b[:, :, 1])
# [[1 2]
# [1 2]]
print(b[:, :, 2])
# [[1 2]
# [1 2]]
Having said that, you can often avoid repeating your arrays altogether by using broadcasting. For example, let's say I wanted to add a (3,) vector:
c = np.array([1, 2, 3])
to a. I could copy the contents of a 3 times in the third dimension, then copy the contents of c twice in both the first and second dimensions, so that both of my arrays were (2, 2, 3), then compute their sum. However, it's much simpler and quicker to do this:
d = a[..., None] + c[None, None, :]
Here, a[..., None] has shape (2, 2, 1) and c[None, None, :] has shape (1, 1, 3)*. When I compute the sum, the result gets 'broadcast' out along the dimensions of size 1, giving me a result of shape (2, 2, 3):
print(d.shape)
# (2, 2, 3)
print(d[..., 0]) # a + c[0]
# [[2 3]
# [2 3]]
print(d[..., 1]) # a + c[1]
# [[3 4]
# [3 4]]
print(d[..., 2]) # a + c[2]
# [[4 5]
# [4 5]]
Broadcasting is a very powerful technique because it avoids the additional overhead involved in creating repeated copies of your input arrays in memory.
* Although I included them for clarity, the None indices into c aren't actually necessary - you could also do a[..., None] + c, i.e. broadcast a (2, 2, 1) array against a (3,) array. This is because if one of the arrays has fewer dimensions than the other then only the trailing dimensions of the two arrays need to be compatible. To give a more complicated example:
a = np.ones((6, 1, 4, 3, 1)) # 6 x 1 x 4 x 3 x 1
b = np.ones((5, 1, 3, 2)) # 5 x 1 x 3 x 2
result = a + b # 6 x 5 x 4 x 3 x 2
Another way is to use numpy.dstack. Supposing that you want to repeat the matrix a num_repeats times:
import numpy as np
b = np.dstack([a]*num_repeats)
The trick is to wrap the matrix a into a list of a single element, then using the * operator to duplicate the elements in this list num_repeats times.
For example, if:
a = np.array([[1, 2], [1, 2]])
num_repeats = 5
This repeats the array of [1 2; 1 2] 5 times in the third dimension. To verify (in IPython):
In [110]: import numpy as np
In [111]: num_repeats = 5
In [112]: a = np.array([[1, 2], [1, 2]])
In [113]: b = np.dstack([a]*num_repeats)
In [114]: b[:,:,0]
Out[114]:
array([[1, 2],
[1, 2]])
In [115]: b[:,:,1]
Out[115]:
array([[1, 2],
[1, 2]])
In [116]: b[:,:,2]
Out[116]:
array([[1, 2],
[1, 2]])
In [117]: b[:,:,3]
Out[117]:
array([[1, 2],
[1, 2]])
In [118]: b[:,:,4]
Out[118]:
array([[1, 2],
[1, 2]])
In [119]: b.shape
Out[119]: (2, 2, 5)
At the end we can see that the shape of the matrix is 2 x 2, with 5 slices in the third dimension.
Use a view and get free runtime! Extend generic n-dim arrays to n+1-dim
Introduced in NumPy 1.10.0, we can leverage numpy.broadcast_to to simply generate a 3D view into the 2D input array. The benefit would be no extra memory overhead and virtually free runtime. This would be essential in cases where the arrays are big and we are okay to work with views. Also, this would work with generic n-dim cases.
I would use the word stack in place of copy, as readers might confuse it with the copying of arrays that creates memory copies.
Stack along first axis
If we want to stack input arr along the first axis, the solution with np.broadcast_to to create 3D view would be -
np.broadcast_to(arr,(3,)+arr.shape) # N = 3 here
Stack along third/last axis
To stack input arr along the third axis, the solution to create 3D view would be -
np.broadcast_to(arr[...,None],arr.shape+(3,))
If we actually need a memory copy, we can always append .copy() there. Hence, the solutions would be -
np.broadcast_to(arr,(3,)+arr.shape).copy()
np.broadcast_to(arr[...,None],arr.shape+(3,)).copy()
Here's how the stacking works for the two cases, shown with their shape information for a sample case -
# Create a sample input array of shape (4,5)
In [55]: arr = np.random.rand(4,5)
# Stack along first axis
In [56]: np.broadcast_to(arr,(3,)+arr.shape).shape
Out[56]: (3, 4, 5)
# Stack along third axis
In [57]: np.broadcast_to(arr[...,None],arr.shape+(3,)).shape
Out[57]: (4, 5, 3)
Same solution(s) would work to extend a n-dim input to n+1-dim view output along the first and last axes. Let's explore some higher dim cases -
3D input case :
In [58]: arr = np.random.rand(4,5,6)
# Stack along first axis
In [59]: np.broadcast_to(arr,(3,)+arr.shape).shape
Out[59]: (3, 4, 5, 6)
# Stack along last axis
In [60]: np.broadcast_to(arr[...,None],arr.shape+(3,)).shape
Out[60]: (4, 5, 6, 3)
4D input case :
In [61]: arr = np.random.rand(4,5,6,7)
# Stack along first axis
In [62]: np.broadcast_to(arr,(3,)+arr.shape).shape
Out[62]: (3, 4, 5, 6, 7)
# Stack along last axis
In [63]: np.broadcast_to(arr[...,None],arr.shape+(3,)).shape
Out[63]: (4, 5, 6, 7, 3)
and so on.
Timings
Let's use a large sample 2D case and get the timings and verify output being a view.
# Sample input array
In [19]: arr = np.random.rand(1000,1000)
Let's prove that the proposed solution is a view indeed. We will use stacking along first axis (results would be very similar for stacking along the third axis) -
In [22]: np.shares_memory(arr, np.broadcast_to(arr,(3,)+arr.shape))
Out[22]: True
Let's get the timings to show that it's virtually free -
In [20]: %timeit np.broadcast_to(arr,(3,)+arr.shape)
100000 loops, best of 3: 3.56 µs per loop
In [21]: %timeit np.broadcast_to(arr,(3000,)+arr.shape)
100000 loops, best of 3: 3.51 µs per loop
Being a view, increasing N from 3 to 3000 changed nothing on timings and both are negligible on timing units. Hence, efficient both on memory and performance!
This can now also be achived using np.tile as follows:
import numpy as np
a = np.array([[1,2],[1,2]])
b = np.tile(a,(3, 1,1))
b.shape
(3,2,2)
b
array([[[1, 2],
[1, 2]],
[[1, 2],
[1, 2]],
[[1, 2],
[1, 2]]])
A=np.array([[1,2],[3,4]])
B=np.asarray([A]*N)
Edit #Mr.F, to preserve dimension order:
B=B.T
Here's a broadcasting example that does exactly what was requested.
a = np.array([[1, 2], [1, 2]])
a=a[:,:,None]
b=np.array([1]*5)[None,None,:]
Then b*a is the desired result and (b*a)[:,:,0] produces array([[1, 2],[1, 2]]), which is the original a, as does (b*a)[:,:,1], etc.
Summarizing the solutions above:
a = np.arange(9).reshape(3,-1)
b = np.repeat(a[:, :, np.newaxis], 5, axis=2)
c = np.dstack([a]*5)
d = np.tile(a, [5,1,1])
e = np.array([a]*5)
f = np.repeat(a[np.newaxis, :, :], 5, axis=0) # np.repeat again
print('b='+ str(b.shape), b[:,:,-1].tolist())
print('c='+ str(c.shape),c[:,:,-1].tolist())
print('d='+ str(d.shape),d[-1,:,:].tolist())
print('e='+ str(e.shape),e[-1,:,:].tolist())
print('f='+ str(f.shape),f[-1,:,:].tolist())
b=(3, 3, 5) [[0, 1, 2], [3, 4, 5], [6, 7, 8]]
c=(3, 3, 5) [[0, 1, 2], [3, 4, 5], [6, 7, 8]]
d=(5, 3, 3) [[0, 1, 2], [3, 4, 5], [6, 7, 8]]
e=(5, 3, 3) [[0, 1, 2], [3, 4, 5], [6, 7, 8]]
f=(5, 3, 3) [[0, 1, 2], [3, 4, 5], [6, 7, 8]]
Good luck
I'm trying to get the indices to sort a multidimensional array by the last axis, e.g.
>>> a = np.array([[3,1,2],[8,9,2]])
And I'd like indices i such that,
>>> a[i]
array([[1, 2, 3],
[2, 8, 9]])
Based on the documentation of numpy.argsort I thought it should do this, but I'm getting the error:
>>> a[np.argsort(a)]
IndexError: index 2 is out of bounds for axis 0 with size 2
Edit: I need to rearrange other arrays of the same shape (e.g. an array b such that a.shape == b.shape) in the same way... so that
>>> b = np.array([[0,5,4],[3,9,1]])
>>> b[i]
array([[5,4,0],
[9,3,1]])
Solution:
>>> a[np.arange(np.shape(a)[0])[:,np.newaxis], np.argsort(a)]
array([[1, 2, 3],
[2, 8, 9]])
You got it right, though I wouldn't describe it as cheating the indexing.
Maybe this will help make it clearer:
In [544]: i=np.argsort(a,axis=1)
In [545]: i
Out[545]:
array([[1, 2, 0],
[2, 0, 1]])
i is the order that we want, for each row. That is:
In [546]: a[0, i[0,:]]
Out[546]: array([1, 2, 3])
In [547]: a[1, i[1,:]]
Out[547]: array([2, 8, 9])
To do both indexing steps at once, we have to use a 'column' index for the 1st dimension.
In [548]: a[[[0],[1]],i]
Out[548]:
array([[1, 2, 3],
[2, 8, 9]])
Another array that could be paired with i is:
In [560]: j=np.array([[0,0,0],[1,1,1]])
In [561]: j
Out[561]:
array([[0, 0, 0],
[1, 1, 1]])
In [562]: a[j,i]
Out[562]:
array([[1, 2, 3],
[2, 8, 9]])
If i identifies the column for each element, then j specifies the row for each element. The [[0],[1]] column array works just as well because it can be broadcasted against i.
I think of
np.array([[0],
[1]])
as 'short hand' for j. Together they define the source row and column of each element of the new array. They work together, not sequentially.
The full mapping from a to the new array is:
[a[0,1] a[0,2] a[0,0]
a[1,2] a[1,0] a[1,1]]
def foo(a):
i = np.argsort(a, axis=1)
return (np.arange(a.shape[0])[:,None], i)
In [61]: foo(a)
Out[61]:
(array([[0],
[1]]), array([[1, 2, 0],
[2, 0, 1]], dtype=int32))
In [62]: a[foo(a)]
Out[62]:
array([[1, 2, 3],
[2, 8, 9]])
The above answers are now a bit outdated, since new functionality was added in numpy 1.15 to make it simpler; take_along_axis (https://docs.scipy.org/doc/numpy-1.15.1/reference/generated/numpy.take_along_axis.html) allows you to do:
>>> a = np.array([[3,1,2],[8,9,2]])
>>> np.take_along_axis(a, a.argsort(axis=-1), axis=-1)
array([[1 2 3]
[2 8 9]])
I found the answer here, with someone having the same problem. They key is just cheating the indexing to work properly...
>>> a[np.arange(np.shape(a)[0])[:,np.newaxis], np.argsort(a)]
array([[1, 2, 3],
[2, 8, 9]])
You can also use linear indexing, which might be better with performance, like so -
M,N = a.shape
out = b.ravel()[a.argsort(1)+(np.arange(M)[:,None]*N)]
So, a.argsort(1)+(np.arange(M)[:,None]*N) basically are the linear indices that are used to map b to get the desired sorted output for b. The same linear indices could also be used on a for getting the sorted output for a.
Sample run -
In [23]: a = np.array([[3,1,2],[8,9,2]])
In [24]: b = np.array([[0,5,4],[3,9,1]])
In [25]: M,N = a.shape
In [26]: b.ravel()[a.argsort(1)+(np.arange(M)[:,None]*N)]
Out[26]:
array([[5, 4, 0],
[1, 3, 9]])
Rumtime tests -
In [27]: a = np.random.rand(1000,1000)
In [28]: b = np.random.rand(1000,1000)
In [29]: M,N = a.shape
In [30]: %timeit b[np.arange(np.shape(a)[0])[:,np.newaxis], np.argsort(a)]
10 loops, best of 3: 133 ms per loop
In [31]: %timeit b.ravel()[a.argsort(1)+(np.arange(M)[:,None]*N)]
10 loops, best of 3: 96.7 ms per loop