How to add value at end of row in a matrix

It is a very basic question but I could not figure how to do it without a for loop.

I have two vectors R and C of same length

R'=[2 4 5 2]
C'=[1 3 6 7]

I want to have

A= 
[0 0
 1 7
 0 0
 3 0
 6 0]

Basically R is for the rows and I want to have the content of C in the corresponding row in matrix A

If I do

    A(R)=C

it is going to be a vector not a matrix and the newest value will erase the oldest value in each row

A(R,:) will assign the value to the whole line which is not also what I want.

how can I increase the index of the column if there is more than one entry for each row?

[Muhammad Usman Saleem]

"Basically R is for the rows and I want to have the content of C in the corresponding row in matrix A"

What is meany by R is for rows, is this means you want A matrix rows as no of rows in R? and remaining portion of line is totally unable to understand ,plz

Answer 1

[Stephen23]

Without any loops:

R = [2;4;5;2];
C = [1;3;6;7];
[~,X] = sort(R);
[~,~,Y] = unique(R);
Z = accumarray(Y,ones(size(R)),[],@(v){v});
Y(X) = cell2mat(cellfun(@cumsum,Z,'UniformOutput',false));
A = zeros(max(R),max(Y));
A(sub2ind(size(A),R,Y)) = C

creates this output:

A =
   0   0
   1   7
   0   0
   3   0
   6   0

[Muhammad Usman Saleem]

can some one tell me what i want to know in the above commments?

[etudiant_is]

It does what I want but can you please explain to me the lines for Z and Y(X). I don't understand yet the use of @ in matlab. Thanks

[Stephen23]

@etudiant_is: here is the code with comments to help you:

% sort the values of R:
[~,X] = sort(R);
% get indices/groups of unique values in R:
[~,~,Y] = unique(R);
% collect each group's values into a vector in a cell array:
Z = accumarray(Y,ones(size(R)),[],@(v){v});
% create consecutive column indices for each value in each group:
Y(X) = cell2mat(cellfun(@cumsum,Z,'UniformOutput',false));
% preallocate an output matrix of the correct size:
A = zeros(max(R),max(Y));
% use linear indices to allocate the original values into the output matrix:
A(sub2ind(size(A),R,Y)) = C

And in particular the lines that you ask about. The vector Y contains a list of unique groups/rows. For your example this is

>> [~,~,Y] = unique(R)
Y =
     1
     2
     3
     1

We use these group indices to collect some 1's together in a cellarray, they will be later converted into the column indices:

>> Z = accumarray(Y,ones(size(R)),[],@(v){v});
>> Z{:}
ans =
     1
     1
ans =
     1
ans =
     1

To convert these vectors of 1's into consecutive column indices we use cumsum, which has to be encapsulated in a cellfun because the groups of ones are held in a cell array:

>> tmp = cellfun(@cumsum,Z,'UniformOutput',false);
>> tmp{:}
ans =
     1
     2
ans =
     1
ans =
     1

These are joined together using cell2mat to give one numeric vector:

>> cell2mat(tmp)
ans =
     1
     2
     1
     1

Now we have to right column indices, but in the wrong order. We can use the sort index X to put them back into the correct order:

>> Y(X) = cell2mat(tmp)
Y =
     1
     1
     1
     2

I reused the variable Y because this allow us to keep its shape (column vector), otherwise a reshape would be required. Now we have a list of row indices R and column indices Y. These can be used with sub2ind to get the linear indices, and these then used to allocate the value allocated to the output matrix.

The @ is used to define a function handle, which allows the function to be passed around like a variable. This is quite handy!

[etudiant_is]

Thanks for the detailed explanation. I will take my time to understand it.

Answer 2

[Andrei Bobrov]

R=[2 4 5 2]
C=[1 3 6 7]

[b,ii] = histc(R(:),unique(R));
out = zeros(max(R),max(b));
for  jj = 1:numel(b)
    out(R(find(jj==ii,1,'first')),1:b(jj)) = C(jj==ii);
end

[etudiant_is]

Thanks for your answer, but why do you take only first value in find?

Answer 3

[Muhammad Usman Saleem]

I unable to understand what you want in your final matrix.

First see if you wanting column vectors these are not correct format in matlab, see it below

>> R'=[2 4 5 2]
C'=[1 3 6 7]

output

 R'=[2 4 5 2]
   |
Error: The expression to the left of the equals sign is not a
valid target for an assignment.

Now try this if you want column vectors

>> R=[2 4 5 2]'

R =

       2
       4
       5
       2

>> C=[1 3 6 7]'

C =

       1
       3
       6
       7

Now

>> A(R)=C

A =

       0     7     0     3     6

[etudiant_is]

For the transpose, I just wanted to write the vector here horizontally to be more clear, that is all.

What I want is to take the number of the rows from vector R and then see what is in C and put it in these corresponding rows. In my example, I have

R(1)=2 and C(1)=1

so I am going to have

A(2, 1)=1

then at the end of R, I have

R(4)=2 and 

C(4)=7 again

so I will have

A(2,2)= 7 and so on.

Hope it is clearer.