How to find a value in a list?
Mostra commenti meno recenti
I have a m-by-3 matrix of numbers (as below) and I would like to find a specific number in a row and then find the other numbers in that same row then write these in a seperate matrix. For example I want to look for the numbers connected to the number 4, in row 1 these will be 15 and 12 and in row 15 these will be 12 and 3. I then want a matrix called 4 to contain the values 15, 12 and 3. I do not want any repeats in this new matrix, hence 12 only appears once in the new matrix.
15 4 12
5 7 2
22 23 21
18 11 10
10 8 14
16 12 11
12 3 6
8 7 5
8 5 9
14 8 9
7 1 2
5 13 9
5 2 13
7 6 1
12 4 3
6 Commenti
Thomas
il 15 Ago 2012
This looks like a Homework question. What have you done so far? Where are you getting stuck..
Thomas
il 15 Ago 2012
@ Illam and Andrei: I thought HWk questions would not be answered on the forum without the poster showing what he/she has done...
Jordan
il 15 Ago 2012
Thomas
il 15 Ago 2012
Jordan, sorry If my comment offended you, but this is the kind of problem I usually give the class I teach for introduction to Matlab and we do not want the forum to be a Hwk solving forum..
Matt Fig
il 15 Ago 2012
Questions:
Do you want all of the numbers in a row that has the given number in it, or just those connected to it (adjacent neighbors)? Take the number 15 in your given list. Do you want to include both the 4 and the 12 from row 1 or just the 4?
What if we had a row like this:
4 4 6
Would we count 4 as a neighbor to 4 (Andrei's code will not)?
Will the real array you need to work with have numbers only on [1,23] or what range?
Thanks.
Jordan
il 16 Ago 2012
Risposte (4)
Ilham Hardy
il 15 Ago 2012
Unchecked..
test_val = [15 4 12;5 7 2;22 23 21;18 11 10;10 8 14;16 12 11;12 3 6;8 7 5;8 5 9;14 8 9;7 1 2;5 13 9;5 2 13;7 6 1;12 4 3];
base_num = 4;
[r4 c4] = find(test_val==base_num);
matrx4 = unique(test_val(r4,:));
matrx4(matrx4==base_num)=[];
HTH, IH
Andrei Bobrov
il 15 Ago 2012
Modificato: Andrei Bobrov
il 16 Ago 2012
z = [...
15 4 12
5 7 2
22 23 21
18 11 10
10 8 14
16 12 11
12 3 6
8 7 5
8 5 9
14 8 9
7 1 2
5 13 9
5 2 13
7 6 1
12 4 3];
out = setdiff(unique(z(any(z == 4,2),:)),4);
ADD ( EDIT )
c = unique(z);
n = numel(c);
out = cell(1,n);
for j1 = 1:n
out{j1} = unique(z(conv2(+(z==c(j1)),[1 0 1],'same')>0));
end
Jordan
il 15 Ago 2012
0 voti
4 Commenti
Andrei Bobrov
il 15 Ago 2012
See ADD in my answer.
Matt Fig
il 15 Ago 2012
Also, please look at my questions in the comment section. I want to be sure you get the correct approach...
Jordan
il 16 Ago 2012
Andrei Bobrov
il 16 Ago 2012
Hi Jordan! Please use:
c = unique(z);
n = numel(c);
out = cell(1,n);
for j1 = 1:n
out{j1} = unique(z(conv2(+(z==c(j1)),[1 0 1],'same')>0));
end
Matt Fig
il 16 Ago 2012
For large data sets, this is the fastest I could come up with:
A = ceil(rand(19e4,3)*255); % Large data set on [1 255]
% Begin Code, store results in cell array T.
m = max(A(:));
T = cell(1,m);
for ii = 1:m
T{ii} = A(any(A==ii,2),:);
T{ii} = sort(T{ii}(:).');
T{ii} = T{ii}([~isempty(T{ii}) diff(T{ii})~=0]);
T{ii} = T{ii}(T{ii}~=ii);
end
Categorie
Scopri di più su Matrix Indexing 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!