Yes, definitely.
filename = 'AppropriateFileName.csv'
T = readtable(filename);
G = findgroups(T{:,1}); %first column
Tc = splitapply( @(varargin) varargin, T, G);
Now Tc is a cell array of table objects, one cell entry for each unique identifier from the first column.
I use a bit of a hack here to do the cell array conversion. When you call a function with arguments, and the function uses the special parameter name varargin, then within the function you can refer to the entire set of arguments from that position onwards, as the cell array varargin, with the first entry being the first parameter there, the second being the second parameter there, and so on. within the @(varargin) anonymous function, whatever was passed in as potentially multiple arguments can be referred to as the cell varargin. And what I return from the function is varargin -- which is to say the cell array of arguments. The effect is the same as
@(var1, var2, var3, ... varN) {var1, var2, var3, ..., varN}
but you do not need to worry about how many arguments there were.
The net result is to wrap each of the partitioned table sections into a cell.
