Create new table wit differentiate or diff from existing timetable

I've got below timetable with daily measurements from my gas meter at home:

Timestamps GasM3

____________________ _____________

14-Aug-2019 00:00:00 331.15

25-Aug-2019 00:00:00 332.37

26-Aug-2019 00:00:00 332.74

27-Aug-2019 00:00:00 333.97

I would like to know the daily consumption, so I tried playing with "differentiate" & "diff" to get these values in a new table.

I've tried

Test1 = differentiate(DailyGasDataTimeTable)

which gave the following error:

Error using Visualization: Differential test</a> (line 17</a>)

'differentiate' requires Curve Fitting Toolbox

and

Test2 = diff(DailyGasDataTimeTable)

which gave the following error:

Undefined function 'diff' for input arguments of type 'table'.

Error in Visualization: Differential test (line 17)

Unfortunately I'm stuck, please advise.

Kind regards,

Daniel

Answer 1

[Vinod]

You can extract the specific column, GasM3, from the timetable and use the diff function. Something like this:

data = thingSpeakRead(YourChannelNum,'OutputFormat','TimeTable');
deltaM3 = diff(data.GasM3);
newTable = timetable(data.Timestamps(2:end), deltaM3)

[Daniel Ubbink]

Dear Vinod,

Great, it works!

Thank you very much

Regards Daniel

[Walter Roberson]

Note that it might make more sense to use gradient than diff

Answer 2

[Walter Roberson]

Use a timetable() object and use retime()

Answer 3

[Kristina Collins]

What if my timetable is missing entries, or the entries are unevenly timed?

[Walter Roberson]

Use retime(), See also fillmissing()

Answer 4

[Olavo]

Hello, Thanks for your answer it's very helpful!

What if I the variable I what to differentiate is not a fixed name and will be changing along the code? for example:

for i=1:10;
    deltaM(i) = diff(data.GasM3{i}); % -->> does not work with this notation
end

[Walter Roberson]

N = 10;
deltaM = cell(N,1);
for i = 1 : N
    varname = "GasM3" + i;
    deltaM{i} = diff(data.(varname));
end

This assumes you are looking for GasM31 GasM32 GasM33 and so on up to GasM310

[Steven Lord]

Or you could use a numeric index.

T = array2table(magic(4))

T = 4×4 table
    Var1    Var2    Var3    Var4
    ____    ____    ____    ____

     16       2       3      13 
      5      11      10       8 
      9       7       6      12 
      4      14      15       1 

A = T{:, 3} % Equivalent to T.Var3

A = 4×1

3 10 6 15

If you're release R2023a or later, you may be able to compute directly on the table or timetable, as described on this documentation page.

T2 = diff(T)

T2 = 3×4 table
    Var1    Var2    Var3    Var4
    ____    ____    ____    ____

    -11       9       7      -5 
      4      -4      -4       4 
     -5       7       9     -11 

[Olavo]

Thanks Walter and Steven, very useful information.

Just to add some extra information, I found the function wildcardPattern very useful to get similar names from table. would be something like:

varname = "GasM3" + wildcardPattern;

Matches as few characters of any type - MATLAB wildcardPattern - MathWorks Nordic