Sort 2D list based on how I sort another 2D list rows with Python

Question

I have a 2D list having many elements and row of different length. Let's call it ListA. I order listA like this:

listA.sort(key=len)

I have another 2D list (listB) which I would like to sort like A is sorted. If row 3 of A becomes the first row, the third row of B must go first and so on.

How can I do?

I don't want to use numpy.

EDIT: I try to be more clear:

Suppose a matrix A made like this (The original matrix is a lot bigger)

A = [
    [
        [86,98,98,0,0]
    ],

    [
        [79,71,105,1,1],  [79,71,106,1,1],  [80,72,105,0,2]
    ], 

    [
        [86,81,27,1,1],   [85,80,25,1,0]
    ],

    [
        [99,80,73,1,1],  [99,81,73,2,1]
    ]

]

this matrix has 4 rows with different length (1, 3, 2, 2). Actually each length contains the coordinates and other values of my analysis. I want to order A so that first I have the shortest row and at the end the longest one. In this case (1, 2, 2, 3)

Now I also have another matrix B which has the same number of rows as A but might have different length compared to A.

B = [
    [
        [8,79,3],[8,77,42]
    ],
    [
        [10,83,70]
    ],
    [
        [9,81,74],  [13,67,43],  [4,15,88]
    ],
    [
        [5,14,88]
    ]
]

I want to sort the rows of B to correspond the sorted rows in A.

Possible duplicate of [Sorting list based on values from another list?](http://stackoverflow.com/questions/6618515/sorting-list-based-on-values-from-another-list) — Prune, Jan 24 '17 at 17:15
Does A have exactly the same amount of rows of the same length as B? if so, then maybe what you want is a dictionary that takes keys from A and assigns them values of B. Then you can do whatever you need with it. — 098799, Jan 24 '17 at 17:19
Are you not posting an example of the lists to make this just a bit more interesting or what is the idea behind it? — Ma0, Jan 24 '17 at 17:20
None of the answers actually works for me. Maybe I've not been very clear. I'm editing the question in order to make it clearer. All the answer proposed sort the matrix but not in the way I want. With the proposed answers I get the second matrix with the first row containing all the elements of the matrix and sorted while all the other rows are empy. — Michele Paoletti, Jan 31 '17 at 14:38

Ma0 · Answer 1 · 2017-01-24T17:19:27.257

0

This is how i would go about doing it:

a = ['easd', 'sdvgweag', '21x', 'npasp oopp agt']
b = [42, 7642, 2146, 12]

temp = list(zip(a, range(len(a))))
temp.sort(key=lambda x: len(x[0]))
a.sort(key=len)
print(a)  # ['21x', 'easd', 'sdvgweag', 'npasp oopp agt']
print([b[i] for i in [x[1] for x in temp]])  # [2146, 42, 7642, 12]

The idea behind this is to add some mechanism of tracking the changes in list a. This is what zip() does.

edited Jan 24 '17 at 17:19

answered Jan 24 '17 at 17:17

Ma0

15,057
4
35
65

@leaf I didn't notice and thank's for pointing that out but the idea and implementation is exactly the same. I will update when OP posts some examples.. – Ma0 Jan 24 '17 at 17:21

Kaveen Perera · Accepted Answer · 2017-02-01T10:48:56.913

0

With the updated question your problem is much clear now.

You can use a similar logic to my previous answer.

A=[[[86,98,98,0,0]],[[79,71,105,1,1],[79,71,106,1,1],[80,72,105,0,2]], [[86,81,27,1,1],[85,80,25,1,0]], [[99,80,73,1,1],[99,81,73,2,1]]]

B=[[[8,79,3],[8,77,42]],[[10,83,70]],[[9,81,74],[13,67,43],[4,15,88]],[[5,14,88]]]

#check the length of each row and pair it into a new array
tempA =  [] # array
for index in range(len(A)):
    itemLength = len(A[index])
    tempA.append([itemLength,index, A[index]])

tempB =  {} #dict
for index in range(len(B)):
    tempB[index] = B[index]

# sort tempA according to row length
tempA.sort()
B = []
A = []
for item in tempA:
    indexBeforeSorting  = item[1]
    B.append(tempB[indexBeforeSorting])
    A.append( item[2])

This works perfectly

edited Feb 01 '17 at 10:48

answered Jan 24 '17 at 21:01

Kaveen Perera

414
3
12

It does something in a way that create a matrix with the first row containing all the contents ordered and all the others rows empty. But it was not what I needed. Do you have any other ideas? – Michele Paoletti Feb 01 '17 at 09:11
@MichelePaoletti See the updated answer. It works perfectly for your problem. Let me know the results. – Kaveen Perera Feb 01 '17 at 10:52

Sort 2D list based on how I sort another 2D list rows with Python

2 Answers2