Alexis, I am niot 100% sure from your description exactly what you are trying to do. I'm going to assume that you want to create a new dataset that is a vertical concatenation of the first dataset and the second. But since the first has fewer variables, you need to create the missing variables, and you'd like to fill them with NaN.
You mention wanting to insert variables in the first dataset. You can only create a new variables at the end. So if you want to insert new variables in the middle, you need to take two steps: create them at the end, and then use subscripting to rearrange. Given this ...
a = dataset(randn(10,1),randn(10,1),'VarNames',{'W' 'X'});
... you can do something like this ...
a(:,{'Y' 'Z'}) = dataset(NaN(10,1),NaN(10,2));
a = a(:,[1 3:4 2])
... to rearrange. But there may not be any need to do this. Concatenation for dataset works by matching up variable names, and it doesn't matter what order they are in. So you can concatenate these two arrays
a = dataset(randn(10,1),randn(10,1),'VarNames',{'X' 'Y'});
b = dataset(randn(10,1),randn(10,1),'VarNames',{'Y' 'X'});
even though the variables are in a different order.
