Problem when using retime
9 Ansichten (letzte 30 Tage)
Ältere Kommentare anzeigen
Sofie Petersson
am 31 Jan. 2019
Kommentiert: Sofie Petersson
am 1 Feb. 2019
I have a timetable with temperature data from a meteorological station.
I use following function to pick out the daily maximum values and that works fine.
dailymax = retime(temptt, 'daily', 'max');
My problem is that in the column with the date and time (in the the resulting timetable), the time changes to 0:00 in every row. My date format is yyyy-mm-dd hh:mm.
Is there anyone who knows how to fix this?
3 Kommentare
Akzeptierte Antwort
Guillaume
am 31 Jan. 2019
You can't use retime for that. retime is designed to work with timetables that can have more than one variable. In this case, it picks the max of each variable and there's no guarantee the max of each variable is on the same row. And of course, when the aggregration function is something like mean there can't be a matching row.
One way to obtain what you want:
daygroup = discretize(tempTT.Properties.RowTimes, 'day'); %find which group each row belongs to
[~, grouprow] = splitapply(@max, tempTT{:, 2}, daygroup); %find location of max within the group
tablerow = find(diff([0; daygroup])) + grouprow - 1; %add start row of each group to get max location in table
dailymax = tempTT(tablerow, :)
Instead of tempTT.Properties.RowTimes you can use tempTT.NameOfTimeColumn and instead of tempTT{:, 2} you can use tempTT.NameOfVariable
Weitere Antworten (1)
Peter Perkins
am 31 Jan. 2019
Sofie, if I understand correctly, you want to find, within each day, the maximum temperature, and the date/time at which that maximum occurred. Guillaume is correct that retime can't do that, unless you trick it. But there are easier ways.
As Guillaume says, splitapply is one way. Another is in my response to the same question a couple weeks ago.
0 Kommentare
Siehe auch
Kategorien
Mehr zu Tables finden Sie in Help Center und File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!