How replace the duplicated row by the mean

Hi,

I have below data:

1.3
1.5
1.6
1.7
1.7
2.2
2.8
2.8

I want to find the row index of duplicate values (row4 & row5, row row7& row8) and replace these row by mean 1. row4&row5 replace by the average of them, similarly row7&row8

the final output is:

1.3
1.5
1.6
1.7
2.2
2.8

[Guillaume]

The mean of identical values is the same value. So isn't what you're after simply:

unique(yourvector, 'stable')

[Cedric]

Do you have any question regarding my answer to your other question here? :

https://www.mathworks.com/matlabcentral/answers/355855-how-to-read-text-files-from-each-subfolder

Answer 1

[Guillaume]

Assuming that there is more to your question, and that your input is actually a matrix where one of the column has got the indexing values you mention:

[~, ~, group] = unique(yourmatrix(:, idcolumn), 'stable'); 
newmatrix = splitapply(@(rows) mean(rows, 1), yourmatrix, group)

[Mekala balaji]

Sir, I get the following error

??? Undefined function or variable 'idcolumn'.

[Vick]

idcolumn represents the index of the column on which you are having the input data. If your are having the input data(1.3 1.5 1.6 1.7 2.2 2.8) on the first column give 1, if in 3rd column give 3. code will work fine...

[Mekala balaji]

Can it work on 7.10.0.499 (R2010a)?

[Guillaume]

It's a good idea to indicate in your question that you're using an ancient version. No, it can't work in R2010a since splitapply was introduced in R2015b.

This probably works in R2010b:

[~, ~, group] = unique(yourmatrix(:, idcolumn), 'stable'); 
newmatrix = cell2mat(accumarray(group, (1:size(yourmatrix, 1))', [], @(rows) {mean(yourmatrix(rows, :), 1)}))

[Mekala balaji]

still same error: ??? Error using ==> keyboard Undefined function or variable 'idcolumn'.

[Guillaume]

As Surey told you, you need to replace idcolumn by the index of the column you want to use as the ID. I already had to guess that your input was a matrix and not a vector as in your question. I'm afraid I can't read your mind and guess which column is the one with the indexing values.

[Mekala balaji]

Sir, I replaced idcolumn==1,

[~, ~, group] = unique(matrix(:, 1), 'stable'); 
newmatrix = cell2mat(accumarray(group, (1:size(matrix, 1))', [], @(rows) {mean(matrix(rows, :), 1)}))
??? Error using ==> unique at 34
Unrecognized option.

K>> matrix

matrix =

       1     2     3
       2     3     4
       1     2     4

[Guillaume]

Looks like unique didn't have the 'stable' option in 2010. Just remove it. The only difference is that your output matrix will be sorted according to your ID column. If that column is already sorted, it makes no difference.

[Mekala balaji]

Thanks, Sir, I tried in 2015 version and it works.