Devide a table in several tables by criteria

Guys, Hi again,

I didn't want to use my first thread, because this question is a bit different.

I got a timetable which I want to split in several smaller tables using a criteria from one column in the timetable. If you would plot the time and the column Volt you would get several curves. I want to split the data per curve in a single table.

In the timetable Tdata in the column OnOff I got values only 0 or 1.

1 means current flows, 0 no current. So I want that the script goes through the column OnOff and search for the first 1, copy all the rows until the next 0 comes and write an table. And continue this to the end of the column OnOff.

My question is, is something like this possible? And if yes, could you help me with the loop, somehow I have no idea how to solve it.

Many Thanks!

Answer 1

[Peter Perkins]

You can do this in several ways, but honestly the most straight-forward is to write a loop, as shown below. The real question is, why do you want to do this? I'm genuinely asking, because there are tools that let you do things like hourly means without splitting the data up. I guess your answer would be, "I want to plot things separately". Fair enough: you can trick rowfun into doing that but it's a little tricky. Are there other things you want to do on the separate timetables?

Anyways, my solution (which could certainly be tightened up but is at least clear):

>> tt = timetable(rand(15,1),[1;1;1;0;0;1;1;1;1;0;1;1;1;1;1], ...
    'RowTimes',datetime(2019,6,19,0,0,0:14), ...
    'VariableNames',{'X' 'OnOff'});
>> tt.Group = 1 + [0; cumsum(diff([tt.OnOff])~=0)]
tt =
  15×3 timetable
            Time               X        OnOff    Group
    ____________________    ________    _____    _____
    19-Jun-2019 00:00:00     0.69989      1        1  
    19-Jun-2019 00:00:01     0.63853      1        1  
    19-Jun-2019 00:00:02    0.033604      1        1  
    19-Jun-2019 00:00:03    0.068806      0        2  
    19-Jun-2019 00:00:04      0.3196      0        2  
    19-Jun-2019 00:00:05     0.53086      1        3  
    19-Jun-2019 00:00:06     0.65445      1        3  
    19-Jun-2019 00:00:07     0.40762      1        3  
    19-Jun-2019 00:00:08     0.81998      1        3  
    19-Jun-2019 00:00:09     0.71836      0        4  
    19-Jun-2019 00:00:10     0.96865      1        5  
    19-Jun-2019 00:00:11     0.53133      1        5  
    19-Jun-2019 00:00:12     0.32515      1        5  
    19-Jun-2019 00:00:13     0.10563      1        5  
    19-Jun-2019 00:00:14     0.61096      1        5  
>> c = cell(numel(unique(tt.Group)),1);
>> for i = 1:length(c), c{i} = tt(tt.Group == i,1:2); end
>> c{:}
ans =
  3×2 timetable
            Time               X        OnOff
    ____________________    ________    _____
    19-Jun-2019 00:00:00     0.69989      1  
    19-Jun-2019 00:00:01     0.63853      1  
    19-Jun-2019 00:00:02    0.033604      1  
ans =
  2×2 timetable
            Time               X        OnOff
    ____________________    ________    _____
    19-Jun-2019 00:00:03    0.068806      0  
    19-Jun-2019 00:00:04      0.3196      0  
ans =
  4×2 timetable
            Time               X       OnOff
    ____________________    _______    _____
    19-Jun-2019 00:00:05    0.53086      1  
    19-Jun-2019 00:00:06    0.65445      1  
    19-Jun-2019 00:00:07    0.40762      1  
    19-Jun-2019 00:00:08    0.81998      1  
ans =
  1×2 timetable
            Time               X       OnOff
    ____________________    _______    _____
    19-Jun-2019 00:00:09    0.71836      0  
ans =
  5×2 timetable
            Time               X       OnOff
    ____________________    _______    _____
    19-Jun-2019 00:00:10    0.96865      1  
    19-Jun-2019 00:00:11    0.53133      1  
    19-Jun-2019 00:00:12    0.32515      1  
    19-Jun-2019 00:00:13    0.10563      1  
    19-Jun-2019 00:00:14    0.61096      1  

[Stephen23]

@Peter Perkins: +1 well done on very neatly answering the original question.

Of course splitting up a table is unlikely to be required or efficient, as you point out, as tables already support convenient ways to process groups of data:

https://www.mathworks.com/help/matlab/matlab_prog/grouping-variables-for-splitting-data.html

https://www.mathworks.com/help/matlab/matlab_prog/split-table-data-variables-and-apply-functions.html

https://blogs.mathworks.com/loren/2017/08/09/ways-to-perform-calculations-on-groups-of-data-in-tables/

[S.M]

Hi Peter, thank you again for your help. I'll try to adjust it to my needs, thanks a lot.

You guessed it right, I wanted to keep the single curves separated just to plot them a bit easier. But it to me it seems more complicated than I expected :)

[S.M]

Hi Peter, just wanted to say thank you again, everything has worked out really well. Thank you so much!