Concise vector adding in Python?

Question

I often do vector addition of Python lists.

Example: I have two lists like these:

a = [0.0, 1.0, 2.0]
b = [3.0, 4.0, 5.0]

I now want to add b to a to get the result a = [3.0, 5.0, 7.0].

Usually I end up doing like this:

a[0] += b[0]
a[1] += b[1]
a[2] += b[2]

Is there some efficient, standard way to do this with less typing?

UPDATE: It can be assumed that the lists are of length 3 and contain floats.

score 51 · Answer 1 · edited Dec 15 '22 at 19:00

51

If you need efficient vector arithmetic, try Numpy.

>>> import numpy
>>> a=numpy.array([0,1,2])
>>> b=numpy.array([3,4,5])
>>> a+b
array([3, 5, 7])
>>>

Or (thanks, Andrew Jaffe),

>>> a += b
>>> a
array([3, 5, 7])
>>>

edited Dec 15 '22 at 19:00

Glorfindel

21,988
13
81
109

answered May 10 '09 at 11:21

gimel

83,368
10
76
104

nb. OP wants a+=b, really -- see my answer, below! ;-) – Andrew Jaffe May 10 '09 at 11:31
2

Looks like a good way to save typing, but I don't like that it adds a dependency and is three times slower than the original code. – Johan Kotlinski May 10 '09 at 22:29

score 50 · Accepted Answer · answered May 10 '09 at 22:54

I don't think you will find a faster solution than the 3 sums proposed in the question. The advantages of numpy are visible with larger vectors, and also if you need other operators. numpy is specially useful with matrixes, witch are trick to do with python lists.

Still, yet another way to do it :D

In [1]: a = [1,2,3]

In [2]: b = [2,3,4]

In [3]: map(sum, zip(a,b))
Out[3]: [3, 5, 7]

Edit: you can also use the izip from itertools, a generator version of zip

In [5]: from itertools import izip

In [6]: map(sum, izip(a,b))
Out[6]: [3, 5, 7]

Fun fact: `map` comes with `zip` functionality included. This allows to say `map(operator.add, a, b)` — Kos, Nov 05 '17 at 15:03

score 35 · Answer 3 · answered May 10 '09 at 13:15

35

While Numeric is excellent, and list-comprehension solutions OK if you actually wanted to create a new list, I'm surprised nobody suggested the "one obvious way to do it" -- a simple for loop! Best:

for i, bi in enumerate(b): a[i] += bi

Also OK, kinda sorta:

for i in xrange(len(a)): a[i] += b[i]

answered May 10 '09 at 13:15

Alex Martelli

854,459
170
1,222
1,395

+1, this is the obvious solution to me. – Kiv May 10 '09 at 13:47
1

Faster than the other solutions, but still only half as fast. – Johan Kotlinski May 10 '09 at 22:30

Georg Schölly · Answer 4 · 2013-12-19T12:07:54.747

19

If you think Numpy is overkill, this should be really fast, because this code runs in pure C (map() and __add__() are both directly implemented in C):

a = [1.0,2.0,3.0]
b = [4.0,5.0,6.0]

c = map(float.__add__, a, b)

Or alternatively, if you don't know the types in the list:

import operator
c = map(operator.add, a, b)

edited Dec 19 '13 at 12:07

answered May 10 '09 at 17:58

Georg Schölly

124,188
49
220
267

This is a really cool answer, but unfortunately three times slower than the original code. – Johan Kotlinski May 10 '09 at 22:08
1

This really suprises me. At least the first version should be faster. Maybe it's only faster if you have larger lists. – Georg Schölly May 11 '09 at 05:19
3

Have a look at the operator module. You can can just use `operator.add`. – schlamar Dec 19 '13 at 11:29
@schlamar: Changed the answer. Thanks. – Georg Schölly Dec 19 '13 at 12:08
@GeorgSchölly `operator.add` should be even faster – jamylak Feb 23 '15 at 13:48
1

+1 for `operator.add`. Also, in Python 3 `map` produces a map object, not a list. So the correct solution would be `c = list(map(operator.add, a, b))`. – rlat Sep 24 '16 at 21:08

score 12 · Answer 5 · answered May 10 '09 at 11:09

12

How about this:

a = [x+y for x,y in zip(a,b)]

answered May 10 '09 at 11:09

RichieHindle

272,464
47
358
399

2

+1 but this will only work when the lists are of the same size. izip_longest with a zip value of 0 if they are not always of the same size – Nadia Alramli May 10 '09 at 11:14
Nice, but it doesn't save a lot of typing, and the original code is over three times faster. – Johan Kotlinski May 10 '09 at 11:22
@kotlinski- hmm, you are right – Nadia Alramli May 10 '09 at 11:32
what size of input lists do you use to check performance? – SilentGhost May 10 '09 at 11:43
3

who cares about 3? how do you type it for 100 items? – SilentGhost May 10 '09 at 11:49
Well, three is common for representing vertexes in three dimensions :) – Johan Kotlinski May 10 '09 at 11:52
2

well if you want to know how sum two lists of length three pair-wise then you should have explicitly stated that. your method might be the fastest one, but your question is too localized then. – SilentGhost May 10 '09 at 11:58
Sorry, I updated the problem description now. – Johan Kotlinski May 10 '09 at 22:13

score 8 · Answer 6 · edited Dec 15 '22 at 19:00

8

Or, if you're willing to use an external library (and fixed-length arrays), use numpy, which has "+=" and related operations for in-place operations.

import numpy as np
a = np.array([0, 1, 2])
b = np.array([3, 4, 5])
a += b

edited Dec 15 '22 at 19:00

Glorfindel

21,988
13
81
109

answered May 10 '09 at 11:21

Andrew Jaffe

26,554
4
50
59

score 3 · Answer 7 · answered May 10 '09 at 11:11

3

[a[x] + b[x] for x in range(0,len(a))]

answered May 10 '09 at 11:11

ismail

46,010
9
86
95

that's exactly what OP is already doing. – SilentGhost May 10 '09 at 11:12
2

This is the less typing version, please be constructive in your critism. Your blind critism is not helping anyone. – ismail May 10 '09 at 11:15
1

this is neither standard nor efficient. is that constructive enough for you? – SilentGhost May 10 '09 at 11:23
What part of list comprehension & range is not standard enough for you? List comprehensions are efficient, why you think its inefficient? – ismail May 10 '09 at 11:26
3

Chill. Please do something like measure the performance and compare or cite references. – S.Lott May 10 '09 at 11:35
3

about 30 % slower comparing to Nadia's version on my machine. – SilentGhost May 10 '09 at 11:50

score 2 · Answer 8 · answered Sep 02 '15 at 16:41

2

For the general case of having a list of lists you could do something like this:

In [2]: import numpy as np

In [3]: a = np.array([[1, 1, 1], [2, 2, 2], [3, 3, 3],[4, 5, 6]])

In [4]: [sum(a[:,i]) for i in xrange(len(a[0]))]
Out[4]: [10, 11, 12]

answered Sep 02 '15 at 16:41

Ricardo Corral

21
1
2

score 1 · Answer 9 · answered Jun 05 '14 at 20:52

1

If you're after concise, try...

vectors = [[0.0, 1.0, 2.0],[3.0, 4.0, 5.0]]
[sum(col) for col in zip(*vectors)]

Though I can't speak for the performance of this.

answered Jun 05 '14 at 20:52

cefn

2,895
19
28

score 1 · Answer 10 · answered Apr 30 '15 at 23:50

1

list(map(lambda x:x[0]+x[1], zip(a,b)))

answered Apr 30 '15 at 23:50

Victor Carriço

46
5

Can you explain this solution a bit? – dethtron5000 May 01 '15 at 01:40
Sure :) First of all, I applied **zip function** that will return a list of tuples, e.g. if `a=[1,2,3]` and `b=[4,5,6]`, `zip(a,b)` returns `[(1,4), (2,5), (3,6)]`. After that, I applied **map** in this result, **map** function takes a function as first argument and an iterable as second argument and it applies that function to every item of this iterable. In this case, the iterable is `zip(a,b)` and the function is `lambda x: x[0]+x[1]`, it receives an item `x` as argument and it returns `x[0]+x[1]`. `map()` returns a list in python2.In python3 it returns a map object, so I used `list()` – Victor Carriço May 04 '15 at 18:23

score 1 · Answer 11 · edited Aug 10 '16 at 01:33

1

a = map(lambda x, y: x + y, a, b)

edited Aug 10 '16 at 01:33

Martin Tournoij

26,737
24
105
146

answered Aug 09 '16 at 22:01

mwei1us

11
1

score 0 · Answer 12 · answered May 10 '09 at 11:07

0

You could create a function that gets the size of the array, loops through it and creating a return array which it returns.

answered May 10 '09 at 11:07

Ólafur Waage

68,817
22
142
198

Thanks, but that doesn't seem very efficient. – Johan Kotlinski May 10 '09 at 11:10
@kotlinski. What? Size of the array is len(array) which is not a "computation" but an attribute of the list. Looping is certainly less typing than a[0]+=b[0]... a[999]+=b[999]. Also less error-prone. Also more obvious. I'll give up a little efficiency to prevent errors from mistyping. – S.Lott May 10 '09 at 12:51
I think my main gripe would be creating and returning a list, which would not be necessary in this case. Otherwise, yes, it's not worse than the other idea. – Johan Kotlinski May 10 '09 at 22:10
I only said return an array/list is because i thought you wanted that. – Ólafur Waage May 10 '09 at 23:19
"to get the result a = [3.0, 5.0, 7.0]." – Ólafur Waage May 10 '09 at 23:19

score 0 · Answer 13 · answered May 10 '09 at 17:43

An improvement (less memory consumption) of the comprehension list

import itertools a = [x+y for x,y in itertools.izip(a,b)]

Actually if you are not sure that a will be consume then I would even go with generator expression:

(x+y for x,y in itertools.izip(a,b))

Concise vector adding in Python?

13 Answers13

Linked

Related