How do I use movsum with a timetable?

I have a 79880x42 timetable:

Date Station1 Station2 Station3

01-Jan-2011 00:00:00 0 0 0

01-Jan-2011 01:00:00 0.3 0 0

01-Jan-2011 02:00:00 0.4 0 0.6

I want to use movsum function to get the moving sum of 2 hours, 3 hours...12 hours of each station.

So far I have tried M2h = movsum(tt,k,'SamplePoints',Date), where tt=timetable, k=duration (2 hours), Date=datetime vector. But trying different variations of this function only gives me different errors such as "Invalid data type. First input must be numeric or logical." or "Unrecognized function or variable 'Date'.".

Answer 1

[Walter Roberson]

Date = datetime('now') - (days(30):days(-1):days(0)).';
Station1 = sort(rand(31,1));
tt = timetable(Date, Station1)

tt = 31x1 timetable
            Date            Station1 
    ____________________    _________

    23-Jan-2021 22:08:38    0.0039546
    24-Jan-2021 22:08:38     0.050015
    25-Jan-2021 22:08:38      0.10765
    26-Jan-2021 22:08:38      0.17637
    27-Jan-2021 22:08:38      0.19403
    28-Jan-2021 22:08:38      0.25825
    29-Jan-2021 22:08:38      0.27809
    30-Jan-2021 22:08:38      0.27946
    31-Jan-2021 22:08:38      0.31052
    01-Feb-2021 22:08:38      0.31613
    02-Feb-2021 22:08:38      0.34986
    03-Feb-2021 22:08:38      0.41445
    04-Feb-2021 22:08:38      0.41544
    05-Feb-2021 22:08:38       0.4175
    06-Feb-2021 22:08:38      0.50128
    07-Feb-2021 22:08:38      0.53827

k = hours(randi(7))

k = duration
   5 hr

M2h = movsum(tt.Station1, k, 'SamplePoints', tt.Date)

M2h = 31×1

0.0040 0.0500 0.1076 0.1764 0.1940 0.2582 0.2781 0.2795 0.3105 0.3161

[SaaraL]

Thank you for your answer, but it still didn't work for me.

I get two error messages:

1. Error using convertMovfunArgs (line 19)

'SamplePoints' value contains Inf or NaT.

2. Error in datetime/movsum (line 104)

args = convertMovfunArgs(varargin{:});

[Walter Roberson]

find(isnat(tt.Date))