[英]Efficient way to compute distance matrix in NumPy
I have this sample array: 我有这个示例数组:
In [38]: arr
Out[38]: array([ 0, 44, 121, 154, 191])
The above is just a sample whereas my actual array size is pretty huge. 以上只是一个示例,而我的实际数组大小非常大。 So, what is an efficient way to compute a distance matrix?
那么,什么是计算距离矩阵的有效方法?
The result should be: 结果应该是:
In [41]: res
Out[41]:
array([[ 0, 44, 121, 154, 191],
[ -44, 0, 77, 110, 147],
[-121, -77, 0, 33, 70],
[-154, -110, -33, 0, 37],
[-191, -147, -70, -37, 0]])
I wrote a for
loop based implementation which is too slow. 我写了一个基于
for
循环的实现,这个实现太慢了。 Could this be vectorized for efficiency reasons? 出于效率原因,这可以进行矢量化吗?
You can use broadcasting : 你可以使用广播 :
from numpy import array
arr = array([ 0, 44, 121, 154, 191])
arrM = arr.reshape(1, len(arr))
res = arrM - arrM.T
There's subtract
. 有
subtract
。 outer
, which effectively performs broadcasted subtraction between two arrays. outer
,有效地执行两个数组之间的广播减法。
Apply the ufunc
op
to all pairs (a, b) with a in A and b in B.将ufunc
op
应用于所有对(a,b),其中A和B在B中。Let M = A.ndim, N = B.ndim.
设M = A.ndim,N = B.ndim。 Then the result, C, of
op.outer(A, B)
is an array of dimension M + N such that:那么
op.outer(A, B)
的结果C是一个维数为M + N的数组,这样:C[i_0, ..., i_{M-1}, j_0, ..., j_{N-1}] = op(A[i_0, ..., i_{M-1}],B[j_0, ..., j_{N-1}])
np.subtract.outer(arr, arr).T
Or, 要么,
arr - arr[:, None] # essentially the same thing as above
array([[ 0, 44, 121, 154, 191],
[ -44, 0, 77, 110, 147],
[-121, -77, 0, 33, 70],
[-154, -110, -33, 0, 37],
[-191, -147, -70, -37, 0]])
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.