162

I am trying to create a matrix transpose function for python but I can't seem to make it work. Say I have 

theArray = [['a','b','c'],['d','e','f'],['g','h','i']]

and I want my function to come up with

newArray = [['a','d','g'],['b','e','h'],['c', 'f', 'i']]

So in other words, if I were to print this 2D array as columns and rows I would like the rows to turn into columns and columns into rows.

I made this so far but it doesn't work

def matrixTranspose(anArray):
    transposed = [None]*len(anArray[0])
    for t in range(len(anArray)):
        for tt in range(len(anArray[t])):
            transposed[t] = [None]*len(anArray)
            transposed[t][tt] = anArray[tt][t]
    print transposed
Improve this question
0

19 Answers 19

Reset to default
349

Python 2:

>>> theArray = [['a','b','c'],['d','e','f'],['g','h','i']]
>>> zip(*theArray)
[('a', 'd', 'g'), ('b', 'e', 'h'), ('c', 'f', 'i')]

Python 3:

>>> [*zip(*theArray)]
[('a', 'd', 'g'), ('b', 'e', 'h'), ('c', 'f', 'i')]
Improve this answer
Sign up to request clarification or add additional context in comments.

9 Comments

if you're going to iterate through the results, izip from itertools can save memory for large arrays.
How would you have it return a list for the sub lists? Like [['a', 'b', 'g'], ['d', 'e', 'h'], ['c', 'f', 'i']] instead of [('a', 'd', 'g'), ('b', 'e', 'h'), ('c', 'f', 'i')]?
@acollection_: map(list, zip(*theArray)).
@AntonyHatchkins This is not needed with Python 3.0 and above. There, zip already returns an iterator: docs.python.org/3.0/whatsnew/…
@xuiqzy It's not that I'm not aware of it, but that's true.
|
68 
>>> theArray = [['a','b','c'],['d','e','f'],['g','h','i']]
>>> [list(i) for i in zip(*theArray)]
[['a', 'd', 'g'], ['b', 'e', 'h'], ['c', 'f', 'i']]

the list generator creates a new 2d array with list items instead of tuples.

Improve this answer

2 Comments

This is the way to go if you want to assign the result to a variable (as opposed to, e.g., iterating over it directly) — assuming you want lists instead of tuples, as mentioned.
Another option (as implied by the comments in the accepted answer) would be: list(map(list, zip(*theArray)))
42

If your rows are not equal you can also use map:

>>> uneven = [['a','b','c'],['d','e'],['g','h','i']]
>>> map(None,*uneven)
[('a', 'd', 'g'), ('b', 'e', 'h'), ('c', None, 'i')]

Edit: In Python 3 the functionality of map changed, itertools.zip_longest can be used instead:
Source: What’s New In Python 3.0

>>> import itertools
>>> uneven = [['a','b','c'],['d','e'],['g','h','i']]
>>> list(itertools.zip_longest(*uneven))
[('a', 'd', 'g'), ('b', 'e', 'h'), ('c', None, 'i')]
Improve this answer

Comments

21

Much easier with numpy:

>>> arr = np.array([[1,2,3],[4,5,6],[7,8,9]])
>>> arr
array([[1, 2, 3],
       [4, 5, 6],
       [7, 8, 9]])
>>> arr.T
array([[1, 4, 7],
       [2, 5, 8],
       [3, 6, 9]])
>>> theArray = np.array([['a','b','c'],['d','e','f'],['g','h','i']])
>>> theArray 
array([['a', 'b', 'c'],
       ['d', 'e', 'f'],
       ['g', 'h', 'i']], 
      dtype='|S1')
>>> theArray.T
array([['a', 'd', 'g'],
       ['b', 'e', 'h'],
       ['c', 'f', 'i']], 
      dtype='|S1')
Improve this answer

Comments

9

The problem with your original code was that you initialized transpose[t] at every element, rather than just once per row:

def matrixTranspose(anArray):
    transposed = [None]*len(anArray[0])
    for t in range(len(anArray)):
        transposed[t] = [None]*len(anArray)
        for tt in range(len(anArray[t])):
            transposed[t][tt] = anArray[tt][t]
    print transposed

This works, though there are more Pythonic ways to accomplish the same things, including @J.F.'s zip application.

Improve this answer

1 Comment

Note that this implementation doesn't work with matrices that have different numbers of columns and rows
5

To complete J.F. Sebastian's answer, if you have a list of lists with different lengths, check out this great post from ActiveState. In short:

The built-in function zip does a similar job, but truncates the result to the length of the shortest list, so some elements from the original data may be lost afterwards.

To handle list of lists with different lengths, use:

def transposed(lists):
   if not lists: return []
   return map(lambda *row: list(row), *lists)

def transposed2(lists, defval=0):
   if not lists: return []
   return map(lambda *row: [elem or defval for elem in row], *lists)
Improve this answer

2 Comments

That's good catch. However, matrices doesn't have lists with different lengths.
It depends on how they are stored.
5

The "best" answer has already been submitted, but I thought I would add that you can use nested list comprehensions, as seen in the Python Tutorial.

Here is how you could get a transposed array:

def matrixTranspose( matrix ):
    if not matrix: return []
    return [ [ row[ i ] for row in matrix ] for i in range( len( matrix[ 0 ] ) ) ]
Improve this answer

Comments

2

This one will preserve rectangular shape, so that subsequent transposes will get the right result:

import itertools
def transpose(list_of_lists):
  return list(itertools.izip_longest(*list_of_lists,fillvalue=' '))
Improve this answer

Comments

2

you can try this with list comprehension like the following

matrix = [['a','b','c'],['d','e','f'],['g','h','i']] n = len(matrix) transpose = [[row[i] for row in matrix] for i in range(n)] print (transpose)

Improve this answer

Comments

1

If you want to transpose a matrix like A = np.array([[1,2],[3,4]]), then you can simply use A.T, but for a vector like a = [1,2], a.T does not return a transpose! and you need to use a.reshape(-1, 1), as below

import numpy as np
a = np.array([1,2])
print('a.T not transposing Python!\n','a = ',a,'\n','a.T = ', a.T)
print('Transpose of vector a is: \n',a.reshape(-1, 1))

A = np.array([[1,2],[3,4]])
print('Transpose of matrix A is: \n',A.T)
Improve this answer

Comments

1

You may do it simply using python comprehension.

arr = [
    ['a', 'b', 'c'], 
    ['d', 'e', 'f'], 
    ['g', 'h', 'i']
]
transpose = [[arr[y][x] for y in range(len(arr))] for x in range(len(arr[0]))]
Improve this answer

2 Comments

Although this may be a correct answer. Two lines of code isn't very useful without an explanation of what and how it solves the original question. Please provide details to your answer.
when posting a new answer to an old question, expectations are high. Please don't post an inferior solution to the ones already posted
-1 
def matrixTranspose(anArray):
  transposed = [None]*len(anArray[0])

  for i in range(len(transposed)):
    transposed[i] = [None]*len(transposed)

  for t in range(len(anArray)):
    for tt in range(len(anArray[t])):            
        transposed[t][tt] = anArray[tt][t]
  return transposed

theArray = [['a','b','c'],['d','e','f'],['g','h','i']]

print matrixTranspose(theArray)
Improve this answer

Comments

-1 
import  numpy as np #Import Numpy 

m=int(input("Enter row")) #Input Number of row

n=int(input("Enter column")) #Input number of column

a=[] #Blank Matrix

for i in range(m): #Row Input

    b=[] #Blank List

    for j in range(n):#column Input

        j=int(input("Enter Number in Pocket ["+str(i)+"]["+str(j)+"]")) #sow Row Column Number 

        b.append(j) #addVlaue to list

    a.append(b)#Add List To Matrix

a=np.array(a)#convert 1matrix as Numpy

b=a.transpose()#transpose Using Numpy

print(a) #Print Matrix 

print(b)#print Transpose Matrix
Improve this answer

Comments

-2 
#generate matrix
matrix=[]
m=input('enter number of rows, m = ')
n=input('enter number of columns, n = ')
for i in range(m):
    matrix.append([])
    for j in range(n):
        elem=input('enter element: ')
        matrix[i].append(elem)

#print matrix
for i in range(m):
    for j in range(n):
        print matrix[i][j],
    print '\n'

#generate transpose
transpose=[]
for j in range(n):
    transpose.append([])
    for i in range (m):
        ent=matrix[i][j]
        transpose[j].append(ent)

#print transpose
for i in range (n):
    for j in range (m):
        print transpose[i][j],
    print '\n'
Improve this answer

Comments

-3 
a=[]
def showmatrix (a,m,n):
    for i in range (m):
        for j in range (n):
            k=int(input("enter the number")
            a.append(k)      
print (a[i][j]),

print('\t')


def showtranspose(a,m,n):
    for j in range(n):
        for i in range(m):
            print(a[i][j]),
        print('\t')

a=((89,45,50),(130,120,40),(69,79,57),(78,4,8))
print("given matrix of order 4x3 is :")
showmatrix(a,4,3)


print("Transpose matrix is:")
showtranspose(a,4,3)
Improve this answer

Comments

-3 
def transpose(matrix):
   x=0
   trans=[]
   b=len(matrix[0])
   while b!=0:
       trans.append([])
       b-=1
   for list in matrix:
       for element in list:
          trans[x].append(element)
          x+=1
       x=0
   return trans
Improve this answer

Comments

-3 
def transpose(matrix):
    listOfLists = []
    for row in range(len(matrix[0])):
        colList = []
        for col in range(len(matrix)):
            colList.append(matrix[col][row])
    listOfLists.append(colList)

    return listOfLists
Improve this answer

1 Comment

Its a simple implementation for a transpose, though there are libraries like mentioned in other answers are also available.
-3

`

def transpose(m):
    return(list(map(list,list(zip(*m)))))

`This function will return the transpose

Improve this answer

Comments

-3

Python Program to transpose matrix:

row,col = map(int,input().split())
matrix = list()

for i in range(row):
    r = list(map(int,input().split()))
    matrix.append(r)

trans = [[0 for y in range(row)]for x in range(col)]

for i in range(len(matrix[0])):
    for j in range(len(matrix)):
        trans[i][j] = matrix[j][i]     

for i in range(len(trans)):
    for j in range(len(trans[0])):
        print(trans[i][j],end=' ')
    print(' ')
Improve this answer

1 Comment

This is not useful!

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.