how to reduce time grabbing values from indices of cell arrays?

1 visualizzazione (ultimi 30 giorni)
I currently do the operation below and wanted to try to do this without a loop since n is very large.
value = 1:5;
results = zeros(n, 5);
for i = 1:n
idx = ismember(tag{i}, value);
results(i, 1:5) = (data{i}(idx))';
end
n is the number of 1xn cell arrays in the multidimensional one.
value can only be a value that exists in all n tag arrays. In this ex it must be 1 through 5.
results is a nx5 matrix of all the data associated with the value found in tag{n}.
tag and data are both multi dimensional cell arrays where size{tag{n}} == size(data{n}), for example:
tag{1} = {1; 2; 3; 4; 5} data{1} = {34; 42; 43; 52; 1}
tag{2} = {1; 2; 3; 4; 5; 6; 7; 8} data{2} = {32; 09; 12; 33; 4; 98; 67; 11}
tag{3} = {1; 2; 3; 4; 5; 6} data{3} = {98; 28; 48; 29; 9; 22}
tag{n} = {1; 2; 3; 4; 5 ...} data{n} = {x1; x2; x3; x4; x5 ...}
results = [34 42 43 52 1;
32 09 12 33 4;
98 28 48 29 9]
Can this be done with cellfun or similar and save time?

Risposta accettata

Andrei Bobrov
Andrei Bobrov il 7 Nov 2017
d = [data{:}];
out = reshape(d(ismember([tag{:}],1:5)),5,[])';
  4 Commenti
lightroman
lightroman il 7 Nov 2017
hi did you see that tag and data were edited to column vectors? I think just a minor edit fixes, I keep getting dimension mismatches trying to change it.
Andrei Bobrov
Andrei Bobrov il 7 Nov 2017
tag{1} = {1; 2; 3; 4; 5}; data{1} = {34; 42; 43; 52; 1};
tag{2} = {1; 2; 3; 4; 5; 6; 7; 8}; data{2} = {32; 09; 12; 33; 4; 98; 67; 11};
tag{3} = {1; 2; 3; 4; 5; 6}; data{3} = {98; 28; 48; 29; 9; 22};
t = cat(1,tag{:});
t = [t{:}];
d = cat(1,data{:});
d = [d{:}];
out = reshape(d(ismember(t,1:5)),5,[])'

Accedi per commentare.

Più risposte (0)

Categorie

Scopri di più su Structures in Help Center e File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by