# How do you remove duplicates of nested cells?

1 view (last 30 days)
Kris Govertsen on 15 Jun 2019
Commented: Kris Govertsen on 26 Jun 2019
I have cell array that looks like the following:
RomanticGestures = ... % Cell array 5x1
{{List of Flowers};... % nested Cell array 15x1
{List of Chocolates};... % nested Cell array 15x1
{List of Movies};... % nested Cell array 17x1
{List of Flowers};... % nested Duplicate Cell array 15x1
{List of Wines}};... % nested Cell array 34x1
But I do not need the nested cell that contains the list of flowers twice. So this is what I want:
RomanticGestures = ... % Cell array 5x1
{{List of Flowers};... % nested Cell array 15x1
{List of Chocolates;... % nested Cell array 15x1
{List of Movies};... % nested Cell array 17x1
{List of Restaurants}}; % nested Cell array 34x1
I also need to know what row the duplicate occured in.
have tried unique but I think I have the notation wrong. I have tried:
RomanticGestures=unique(RomanticGestures)
Also
RomanticGestures=unique(cell2mat(RomanticGestures),'rows')
cell2mat doesnt work.
Any suggestions are appreciated :) Thank you!

Guillaume on 15 Jun 2019
Sounds like:
Dates{1} = unique(Dates{1}); %What a misleading variable names if it doesn't contain dates/times!
is what you're after.
Kris Govertsen on 26 Jun 2019
This worked! I ended up doing:
[~, idxRomanticGestures] = unique(RomanticGestures);
RomanticGestures=RomanticGestures(idxRomanticGestures);
OtherRelatedArrays=OtherRelatedArrays(idxRomanticGestures);

Jan on 15 Jun 2019
Edited: Jan on 21 Jun 2019
A loop approach:
C = RomanticGestures;
n = numel(C);
remove = false(1, n);
for i1 = 1:n
for i2 = i1 + 1:n
if ~remove(C{i2}) && isequal(C{i1}, C{i2})
remove(i2) = true;
end
end
end
C(remove) = [];
C = RomanticGestures;
n = numel(C);
H = cell(1, n);
for iC = 1:n
H{iC} = DataHash(C{iC}, 'base64');
end
[~, index] = unique(H);
Result = C(index)
##### 2 CommentsShowHide 1 older comment
Jan on 21 Jun 2019
@Kris: Please explain "doesn't work" with any details. I had a typo in my code, which is fixed now, but the problem was the round parentheses for indexing H, instead of the curly braces.
Of course DataHash works for characters.
If you provide some Matlab code, which produces the input data, I could test my code before posting it.