How do I unstack/transpose table in Matlab?

I need to unstack/transpose a table but couldn''t do using unstack ot transpose functions. Please see attached file.

[dpb]

So show us what you tried and what happened...showing an expected output would help I have no idea what you would want that to be.

[MIKHAIL DEREVYANKO]

sorry for complicating things. the real question is as follows

if i have a table of non-numerical variables

id={'id1';'id1';'id2';'id2';'id2'};
cn={'c1';'c2';'c1';'c2';'c3'};
t=table(id,cn);
%unstack/transpose
t1=unstack(t,'cn','id');

the function unstack returns the following error:

Error using tabular/unstack (line 307)

You must specify AGGREGATIONFUNCTION for multiple rows in non-numeric data.

i have no idea what to input for aggregation function.

the desired output should be like this:

id={'id1';'id2'};
cn={'c1','c2',NaN;'c1','c2','c3'};
t1=table(id,cn);
t1=splitvars(t1);

any ideas what should i change/how to achieve the task?

Answer 1

[TADA]

unstack groups using a third column, but you can trick it to do what you want by adding a fake third column

id={'id1';'id1';'id2';'id2';'id2'};
cn={'c1';'c2';'c1';'c2';'c3'};
t=table(id,cn,cn,'VariableNames', {'id', 'cn', 'cn_'});
%unstack/transpose
t1=unstack(t,'cn_','cn');

t1 =

  2×4 table

     id       c1      c2      c3 
    _____    ____    ____    ____

    'id1'    'c1'    'c2'    ''  
    'id2'    'c1'    'c2'    'c3'

[MIKHAIL DEREVYANKO]

Thanks a lot. all works. It's very frustrating that Matlab documentation has nothing on it.

[Guillaume]

That would be because unstack is not designed at all for this use case.

To be honest, I don't see the point of it. The only thing it may help with is visualising the data. Processing it afterward is going to be a lot more complicated than processing what you started with.

[TADA]

I don't know what the OPs use case really is, but sometimes a good visualization is all the processing you need.

I have to agree that in this case theres no reason to blame the documentation, we are tricking unstack to do something it wasn't exatly intended for

[stone_ward]

How were you able to do this without defining an aggregation function? I've been trying to use unstack on my table of non-numeric data, which is similar to the example, except that I do happen to have 3 columns of data. Everytime I try to use it, I get an error about needing and aggregation function. When I supply one, I get the error "AGGREGATIONFUNCTION must return a value whose size conforms to elements of hte data variable 'Data'."

Note: Data is my datavar.

Answer 2

[MIKHAIL DEREVYANKO]

it's nothing to do with the visualisation. I need to pick up correct curves names in a batch validation exercise. it can be done many ways but having a file with aligned discounting curve names for all contracts helps a lot.