changing time step of data or resample the data
显示 更早的评论
I have a motion data with different time step as shown in figure. 1st column represent time step and 2nd to 4th column represent motion. Now, My problem is I want to change the time step of motion data in to uniform time step, say 0.02 sec. How can I change my time step to 0.02 sec time step in matlab? Any answer are highly appreciated. Thank you.
5 个评论
sam moor
2016-11-9
Guillaume
2016-11-11
additional information that was provided as an answer moved here - Guillaume
Here is my data image. 1st column is time and other columns are distance. What I see is the first column is non monotonic that means there are some values which are repeated e.g 1 and 110 rows of first column. so interp1 function is not working. I don't how to solve this repeated problem. I used unique function for column 1 but again the vectors are of different length. So let me know how to solve this problem. Thank you.
Jan
2016-11-11
@sam: Please do not post screenshots. We cannot play with these data or try a suggested solution. Prefer posting data as text or as attached file. Thanks.
sam moor
2016-11-12
sam moor
2016-11-12
回答（2 个）
Sort the rows at first to get an ascending order. Then call interp1:
xx = 0.001:0.02:1.0;
yy = interp1(x, y, xx)
5 个评论
sam moor
2016-11-10
Guillaume
2016-11-10
That's why Jan told you to sort the column.
I don't understand your explanation of the time steps, shouldn't they be
[0.1 0.2 0.3 ... 0.9 1 1.0001 1.0002 .... 59.9999 60] .'
sam moor
2016-11-10
Guillaume
2016-11-11
The problem is not with what it should be, it's not difficult to create that, the problem is with what it is now. Your input is not monotonous, so you need to decide what you want to do about it.
Guillaume
2016-11-10
t = array2table(yourmatrix(:, 2:end), 'rowTimes', ...
datetime(0, 0, 0, 0, 0, yourmatrix(:, 1), 'Format', 'ss.SSS'));
The retime function has then plenty of options to resample your data over the interval and step you want, e.g.:
newtimes = datetime(0, 0, 0, 0, 0, 0) : duration(0, 0, 0.02) : datetime(0, 0, 0, 0, 0, 60);
newtimes.Format = 'ss.SS';
newt = retime(t, newtimes, 'linear');
However, the times in your original array need to be stricly monotonically increasing for any interpolation to work, so you will encounter the same issues as with Jan's answer.
The advantage of a timetable is that there are plenty of built-in function to resample values / aggregate them by period, etc.
3 个评论
Guillaume
2016-11-11
Well, you need to decide what you want to do about your input and these duplicated values. Matlab can't do that for you.
You can indeed only keep one set with unique (but which one, the duplicates are vastly different) but really you need to work out why you have duplicates in the first place and what it means. You could also aggregate the values (average / median / etc.), again only can decide what is appropriate
Once you've worked out what to do, then it is trivial to use interp1 or retime.
sam moor
2016-11-12
类别
在 帮助中心 和 File Exchange 中查找有关 Data Preprocessing 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!
Translated by 
