Recompose stacked column cell vectors into adjacent row cell vectors

I have a row array of cells, varargin, within a function. The contents of each cell corresponds to an argument supplied by the invoker. I know that each argument is a column vector, and the vectors are of the same height. In my particular scenario, they are the input variables of a table when "rowfun" is applied (with "Grouping", though I'm not sure that matters). The type/class of each argument can be anything, including cells, strings, cells containing char arrays, objects of type OptimizationVariable or OptimizationExpression, etc.

I would like to decompose the vertical arrays and recompose them into row cell arrays. For example, if varargin is:

{ [1;2] ["cat";"dog"]}

then I want

{ 1 "cat" ; 2 "dog" }

I know that I can loop through the height of the arguments, then loop through the arguments to buildup each row. I'm looking for vectorized way to do this. If I try [varargin{:}], I get unwanted type conversion, leading to a "3x2 string array":

"1" "cat"
"2" "dog"

Thanks for any suggestion on how to do this while preserving the data type/class. If this is not possible, thanks for letting me know.

Answer 1

[Voss]

Here's one way:

varargin = { [1;2] ["cat";"dog"] };

% convert each element of varargin to a (2x1) cell array:
temp = cellfun(@num2cell,varargin,'UniformOutput',false);

% concatenate the resulting (2x1) cell arrays into a (2x2) cell array:
temp = [temp{:}]

temp = 2×2 cell array
    {[1]}    {["cat"]}
    {[2]}    {["dog"]}

temp{1,1}

ans = 1

temp{1,2}

ans = "cat"

temp{2,1}

ans = 2

temp{2,2}

ans = "dog"

[FM]

Thanks! I should have known that there was a way!

[Voss]

You're welcome!