Strange behavior of "sort" function
Mostra commenti meno recenti
Hello,
The sort function is used as follows:
>> [out,idx] = sort([5 8 91 19])
out =
5 8 19 91
idx =
1 2 4 3
If we need to rearrange the vector later in the code with the previous order, it's possible with that set of figures to write:
>> out(idx)
ans =
5 8 91 19 % is equal to the original vector [5 8 91 19]
All is fine...
But, can you explain why it does'nt work in the same way for the following set :
>> [out,idx] = sort([5 8 4 19])
out =
4 5 8 19
idx =
3 1 2 4
>> out(idx)
ans =
8 4 5 19 % is different from the original vector [5 8 4 19]
I thank you in advance for your answer, I really need to implement that idea of using out(idx) in my code.
Risposta accettata
Più risposte (2)
Ameer Hamza
il 11 Apr 2020
Modificato: Ameer Hamza
il 11 Apr 2020
sort() function does not provide any function to get back the original vector. The first example you gave was just a coincidence. The actual working of idx output of sort is like this
>> A = [5 8 91 19]
A =
5 8 91 19
>> [out,idx] = sort(A)
out =
5 8 19 91
idx =
1 2 4 3
>> A(idx) % <---- idx is used to get sorted vector from origina.
ans =
5 8 19 91
To get the original vector back, you need to call sort twice
A = [5 8 4 19 1];
[out,idx] = sort(A);
[~,idx2] = sort(idx);
out(idx2)
ans =
5 8 4 19 1
2 Commenti
"sort() function does not provide any function to get back the original vector."
Can you please clarify this statement? The two outputs contain enough information to recreate the original vector with one use of basic indexing, nothing else is required.
"To get the original vector back, you need to call sort twice"
No, you do NOT "need" to call sort twice. The two outputs of sort are enough, using basic indexing.
Ameer Hamza
il 11 Apr 2020
Correct. I missed this point. I was pointing the mistake in OPs reasoning and overlooked this simple approach.
Alain LESTRADE
il 11 Apr 2020
0 voti
Categorie
Scopri di più su Shifting and Sorting Matrices in Centro assistenza e File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!