Removing outliers using standard deviation

Hello everyone,

I have a timetable of 8764*3. The first column corresponds to the date, the second to the hour (which goes from 1 to 24 in double format) and the third is a price. My objective is that, for each hour, I remove the prices that are above (Mean of that hour + 3*SD of the prices of that hour) and any price below (Mean of that hour - 3*SD of the prices of that hour). I know I could use the code:

rmoutliers(A,'mean');

However, this filter would take into account all the hours of the sample. Could someone kindly help me to apply it for each hour?

I attach here the data so you can have a clear view of what I have.

Thank you!

Answer 1

[Ive J]

groupfilter does the trick

cleanTable = groupfilter(yourTable, 'Hour', @(x)~isoutlier(x, 'mean'), 'Price');

[Angelavtc]

Thank you @Ive J! And is it possible to identify which date observations for each hour were removed?

[Ive J]

Yes, outTab (complement of cleanTable) would contain outliers per each hour:

outTab = groupfilter(yourTable, 'Hour', @(x)isoutlier(x, 'mean'), 'Price');

[Angelavtc]

Wonderful, thank you @Ive J!