How to sum columns of an array in Python

问题:

How do I add up all of the values of a column in a python array? Ideally I want to do this without importing any additional libraries.

input_val = [[1, 2, 3, 4, 5],
             [1, 2, 3, 4, 5],
             [1, 2, 3, 4, 5]]

output_val = [3, 6, 9, 12, 15]

I know I this can be done in a nested for loop, wondering if there was a better way (like a list comprehension)?

回答1:

zip and sum can get that done:

Code:

[sum(x) for x in zip(*input_val)]

zip takes the contents of the input list and transposes them so that each element of the contained lists is produced at the same time. This allows the sum to see the first elements of each contained list, then next iteration will get the second element of each list, etc...

Test Code:

input_val = [[1, 2, 3, 4, 5],
             [1, 2, 3, 4, 5],
             [1, 2, 3, 4, 5]]

print([sum(x) for x in zip(*input_val)])

Results:

[3, 6, 9, 12, 15]

回答2:

In case you decide to use any library, numpy easily does this:

np.sum(input_val,axis=0)

回答3:

You may also use sum with zip within the map function:

# In Python 3.x 
>>> list(map(sum, zip(*input_val)))
[3, 6, 9, 12, 15]
# explicitly type-cast it to list as map returns generator expression

# In Python 2.x, explicit type-casting to list is not needed as `map` returns list
>>> map(sum, zip(*input_val))
[3, 6, 9, 12, 15]

回答4:

This should work:

[sum(i) for i in zip(*input_val)]

回答5:

Try this:

input_val = [[1, 2, 3, 4, 5],
         [1, 2, 3, 4, 5],
         [1, 2, 3, 4, 5]]

output_val = [sum([i[b] for i in input_val]) for b in range(len(input_val[0]))]

print output_val

回答6:

I think this is the most pythonic way of doing this

map(sum, [x for x in zip(*input_val)])

回答7:

One-liner using list comprehensions: for each column (length of one row), make a list of all the entries in that column, and sum that list.

output_val = [sum([input_val[i][j] for i in range(len(input_val))]) \
                 for j in range(len(input_val[0]))]

回答8:

I guess you can use:

import numpy as np
new_list = sum(map(np.array, input_val))

回答9:

Try this code. This will make output_val end up as [3, 6, 9, 12, 15] given your input_val:

input_val = [[1, 2, 3, 4, 5],
             [1, 2, 3, 4, 5],
             [1, 2, 3, 4, 5]]

vals_length = len(input_val[0])
output_val = [0] * vals_length # init empty output array with 0's
for i in range(vals_length): # iterate for each index in the inputs
    for vals in input_val:
        output_val[i] += vals[i] # add to the same index

print(output_val) # [3, 6, 9, 12, 15]

回答10:

output_val=input_val.sum(axis=0)

this would make the code even simpler I guess

回答11:

Using Numpy you can easily solve this issue in one line:

1: Input

input_val = [[1, 2, 3, 4, 5],
             [1, 2, 3, 4, 5],
             [1, 2, 3, 4, 5]]

2: Numpy does the math for you

np.sum(input_val,axis=0)

3: Then finally the results

array([ 3,  6,  9, 12, 15])