How can I select rows based on the value in a specific column and calculate the mean of those rows for all other columns to create a new table containing just the means?

I have a table array 3152x25 containing data for variables that are monitored over time. In column 25 I have a vaiable 'elapsedHr' for the calculated elapsed time in hours from 0 to 72, with 40+ rows corresponding to each hourly time point. I would like to calculate the mean value of each column variable (columns 2 to 20) for each of the 73 time points and generate a new table containing this data.

[Andrei Bobrov]

Please attach your table array as mat-file or txt-file.

[Elisabeth Slade]

Table attached, Thanks.

Answer 1

[Andrei Bobrov]

TT = table2timetable(Port1a(:,2:end),'RowTime',datetime(Port1a{:,1}));
out = varfun(@mean,TT(:,[1:20,end]),'GroupingVariables','elapsedHr');

or just for your case

out = varfun(@mean,Port1a(:,[2:20,end]),'GroupingVariables','elapsedHr');

[Elisabeth Slade]

Fantastic, thank you so much!

Answer 2

[Peter Perkins]

You could also create a timetable using a duration time vector -- elapsed time from the start instead of absolute times. Equivalent to seconds(Port1a.elapsed, I think. If you did that, you could call retime with 'hourly' to get the hourly means.

Depends what eklse you are doing with the timestamps, of course.