MATLAB Answers

ALDO
0

separate table data in to sub tables

Asked by ALDO
on 19 Apr 2019
Latest activity Answered by Peter Perkins
on 3 May 2019 at 18:47
Hi everyone
I have a table of data the size of 10000x3. There are 300 different names used in the rows. I would like to seperte the variables that have the same name and make a tabel for each variable.
for example if the table below is my original table :
name Variable1 variable2 variable3
A variable1_A1 variable2_A2 variable3_A3
B variable1_B1 variable2_B2 variable3_B3
K variable1_K1 variable2_K2 variable3_K3
L variable1_L1 variable2_L2 variable3_L3
A variable1_A4 variable2_A5 variable3_A6
A variable1_A7 variable2_A8 variable3_A9
L variable1_L4 variable2_L5 variable3_L6
I need to devide it to 4 differnt tables (one for each name):
name Variable1 variable2 variable3
A variable1_A1 variable2_A2 variable3_A3
A variable1_A4 variable2_A5 variable3_A6
A variable1_A7 variable2_A8 variable3_A9
name Variable1 variable2 variable3
L variable1_L1 variable2_L2 variable3_L3
L variable1_L4 variable2_L5 variable3_L6
Thank you in advance for your help!

  0 Comments

Sign in to comment.

2 Answers

Answer by Walter Roberson
on 19 Apr 2019
Edited by Walter Roberson
on 19 Apr 2019
 Accepted Answer

G = findgroups(YourTable.name);
output = splitapply(@(varargin) {table(varargin{:}, 'VariableNames', YourTable.Properties.VariableNames)}, YourTable, G);
This should give a cell array of tables.

  3 Comments

Hi Walter
Thank you for your fast response. could you please explain what c1 c2 c3 c4 represent in splitapply?
Also i tried to run it on my code and got the following error:
Error using splitapply (line 61)
Group numbers must be a vector of positive integers, and cannot be a sparse vector.
I edited the code while you were reading my earlier response.
The c* variable stood in for columns as the documentation implies that the parameters need to be individual variables, but I figured out the way to code around that.
Thank you for taking the time to look over it

Sign in to comment.


Answer by Peter Perkins
on 3 May 2019 at 18:47

ALDO, unless you have a good reason for wanting to do this, I'm gonna suggest that you consider NOT doing it. There are a bunch of functions in MATLAB that are happy to do "grouped" calculations on your original timetable, and you will likely find those easier to use than having to loop over a 300-element cell array of tables.

  0 Comments

Sign in to comment.