Averaging hour value of a timetable

I am new to MatLAB and wanted help. I have hourly precipitation data of a certain place for 30 years (total values/rows=30*365*24=262800). Is there any way i can average values of each hour and generate a new vector having only 12 rows? In other words average of every value located 8640 rows apart.Thanks

[fred ssemwogerere]

Hello, please refer to the link below. It should be of great help.

https://mathworks.com/matlabcentral/answers/504648-multiple-selection-of-an-array

Answer 1

[Akira Agata]

How about the following solution?

% Create sample timetable
Time = datetime(2019,1,1,0,0,0) + hours(0:239)';
Value = rand(240,1);
TT = timetable(Time,Value);

% Calculate average value for each hour
[Group, Hour] = findgroups(TT.Time.Hour);
AvgValue = splitapply(@mean, TT.Value, Group);
tResult = table(Hour,AvgValue);

The result will be like this:

>> tResult

tResult =

24×2 table

Hour AvgValue

____ ________

0 0.61649

1 0.51445

2 0.32941

3 0.60418

4 0.53319

5 0.39934

6 0.41297

7 0.42935

8 0.48803

9 0.50767

10 0.55387

11 0.49949

12 0.59754

13 0.43441

14 0.40604

15 0.40547

16 0.37117

17 0.4985

18 0.5161

19 0.52695

20 0.2493

21 0.66763

22 0.57555

23 0.69638

[Hammad Khan]

Thanks.worked perfect for me

[atmosfera]

Excuse me, How can modify the script If I need to average values of each half hour... Pls

[Eric Beamesderfer]

atmosfera - I had a similar question. I am looking at a full year of data and trust my timeseries (don't need a solution involving a timetable), so here's what I ultimately did for half-hour data (non-leap year):

[Steven Lord]

Use retime with a NEWTIMES input spaced at half an hour intervals and an aggregation method of 'mean'.

newtimes = datetime('now') + hours(0:0.5:6);

If you're doing this outside the context of a timetable, use timeofday to get the elapsed time since midnight and discretize that duration array using a vector of edges created using hours like I did above.