Help ! Average half hourly into 3 hours and save the file as mat with filename reflecting the startTime and Endtime

Hi,

I'm reading half hourly HDF5 files, on which I run some analysis. I want to now average these processed half hourly into 3 hours. This does NOT run.

clc
clear all 
sad = dir( fullfile('path', '*.HDF5' ));  %% all the HDF5 files in the folders for two years and half hourly 
for ii = 1:length(sad)
[data, name] =    read_one_file(sad(ii).name) ; %%% read_one_file where my hour half files are processed 
while (data(ii).TimeID - data(ii).TimeID) == 0  %% TimeID is same for half hourly files within 3 hour windows, so have 8 uniques values for a day
new = nanmean(data(ii).H); %%% here i want to average all the var fields in data strcuture where my TimeID remains same and for same date
extention='.mat';   %%% ext
matname = fullfile('path\test', [[name,num2str(StartTime(ii)),num2str(EndTime(ii))] extention]) % want to save with file with startTime and EndTime
save(matname,'new');
end 
end 

My data structure has following fields; 
data.H ; data.Th ; data.Date ;  data.EndTime ; data.StartTime  ; data.TimeID ;

[Walter Roberson]

What error message do you observe?

The body of your while loop does not change data or ii so if your while loop body is entered, it will ever exit.

Have you considered using timetable() objects with retime() ?

[nlm]

No I haven't. I understand while loop is wrong, I don't know how to move forward...

for ii = 1:length(sad)

if data(ii).TimeID == unique(data(ii).TimeID)

IMERG = nanmean(data(ii).H,3);

end

end

This saves average only for one TimeID. The problem is either I have to run for entire ii which is 17000 HDF5 half an hour files with 2 of 400 * 900 matrices in each file and then average (which I did for 100 files and it works ), or I run for few files and do an 3 hr average and save only that file?

[Walter Roberson]

each iteration of the for loop you are overwriting all of data. You should be storing into data(i) instead . Have it read all of the data first, and then it can make sense to process the data from several files together .