Creating daily average of data set from half hourly data

Question

Sophie Stringer 2019-4-26

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/458720-creating-daily-average-of-data-set-from-half-hourly-data

评论： Sophie Stringer 2019-4-27

Hi there,

I have a data set of variables over 92 days of half hourly recordings. So I have 4,416 data points (48*92 - 48 half hour periods in 1 day)

Instead of having 4,416 data points, I want to create a daily average. So I instead only have 92 averages (or 92 days).

So essential I want to create an average of data points for every 48 data points (1 day)

I want to use a loop to do this, and this is what I have so far, but don't believe its giving me the correct results.

Any help would be much appreciated!

Thank you :D

for i=1:92
    index=i:48:(92*48); 
    data_daily(i)=nanmean(data(index));  
end 

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Adam Danz 2019-4-26

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/458720-creating-daily-average-of-data-set-from-half-hourly-data#answer_372423

编辑：Adam Danz 2019-4-26

在 MATLAB Online 中打开

Option 1

The cleanest solution (IMHO) is to convert your data to a timetable and use retime() to calculate the daily average. No loops needed.

Option 2

If you can't work with timetables and your data are faithfully sampled 48 times per day,

data = rand(4416,1);  %fake data
nDays = 92; 
nSamples = 48; 
dayIdx = repelem((1:nDays)', nSamples, 1); 
dailyAvg = splitapply(@nanmean, data, dayIdx); 

Option 3

If you must use a for-loop for whatever reason and your data are faithfully sampled 48 times per day,

data = rand(4416,1);  %fake data
nDays = 92; 
nSamples = 48; 
dayIdx = repelem((1:nDays)', nSamples, 1); 
dailyAvg = zeros(nDays, 1); 
for i = 1:nDays
    dailyAvg(i) = mean(data(dayIdx==i),'omitnan');    
end

Option 4

If your data are not faithfully sampled 48 times per day, and if you have a vector of time stamps, you can use those time stamps to identify the day and then use findgroups() to create the 'dayIdx' vector in my options 2 and 3.