如何在Python中根据另一个数组中的值提取一个数组中的数据

Question

I have two arrays like:

a = np.array([[0.3, 0.4, 0.3],[0.6, 0.2, 0.2],[0.1, 0.2, 0.7]]) 
b = np.array([[1,2,3], [4,5,6], [7,8,9]])

I hope could get the values in b according to the position of the max value in each row in a, the expected output should be:

[2, 4, 9]

Thanks.

Answer 1

How about:

>>> a = np.array([[0.3, 0.4, 0.3],[0.6, 0.2, 0.2],[0.1, 0.2, 0.7]])
>>> b = np.array([[1,2,3], [4,5,6], [7,8,9]])
>>> b[np.arange(len(b)),a.argmax(axis=1)]
array([2, 4, 9])

Although you should double-check the axis; I always get those backwards.

Answer 2

If you had this as a 2-d python list:

answer = []
for r,row in zip(b,a):
    big = max(enumerate(row), key=operator.itemgetter(1))
    answer.append(r[big[0]])

Of course, you could do this as a one-liner:

answer = [r[max(enumerate(row), key=operator.itemgetter(1))[0]] for r,row in zip(b,a)]

Answer 3

def b_from_a(a,b):
    if len(a) != len(b):
        raise ValueError("Both lists should be the same length")
    for i,element in enumerate(a):
        if len(a[i]) != len(b[i]):
            raise ValueError("The lists in element {} are not of equal length".format(i))
        i_of_max = a.index(max(a))
        yield b[i][i_of_max]

This is in pure python. I don't do any work with NumPy so although I'm sure there's a better bit of code that will get you where you want to go, this should work.

Answer 4

try this:

result = np.zeros(len(a));
for i in range(len(a)):
    result[0] = b[i][a[i].argmax()];

print(result)

(as long as np stands for NumPy)